# Odeist AI: full text > Odeist AI (https://odeist.com) is an AI social media agent. It finds people asking for what you sell in live conversations on X (Twitter) and LinkedIn, replies in your brand voice, and writes and publishes posts with AI images and short videos to X, LinkedIn and Instagram. You approve every reply and post from WhatsApp, or let it run on autopilot. Built for founders, creators, small teams and agencies. Reddit works through Odeist Scout, our app on Reddit, which finds threads asking for what you sell and drafts replies you post yourself. ## Features - Finds your people: Keyword search on X and LinkedIn - Hacker News: Threads, and Show HN drafts at launch - Creator outreach: Finds creators and drafts the pitch you send - Reddit, via Odeist Scout: Our app on Reddit finds threads and drafts replies - Competitors: Weekly watch on their sites and news - Posts in your voice: Text, images, video and carousels, on schedule - Articles: Published to WordPress, Webflow, Sanity and more - UGC clips: For TikTok, Reels and Shorts - Launch kit: Thread, post, email, press release, plan - Daily SEO audit: Ready fixes, Search Console and GA4 - AI answers: Tracks ChatGPT and Gemini - Ads: Meta campaigns, with limits; more coming soon - Ask first, or on its own: You choose, per agent - Approve anywhere: The app, Slack, WhatsApp, Telegram - Morning rundown: Today in the app, and a daily digest - Learns from you: Every skip and lesson sticks - Your AI tools: MCP access, plus Notion, Linear and your CRM - Brands and teammates: Calendar, analytics and reports for each ## Coming soon ## How it works 1. It finds the people asking, and knows when to stay quiet.: Your agents read X and LinkedIn for people asking for what you sell, comparing your competitors or describing the problem you solve. They draft a reply in your voice where it fits, and skip the rest. 2. One brief. A whole team behind it.: Eleven specialists, one brand profile written from your site. Replies, posts, articles, SEO, ChatGPT answers, Hacker News, creators, clips, ads and competitors, all at once, handing work along: an SEO gap becomes the article that fills it. 3. Approve with a tap, until you trust it.: Drafts arrive in WhatsApp, Telegram, Slack or the app. Approve, edit or skip, or just say “shorter” and it remembers. When it sounds like you, X, LinkedIn, Instagram and Articles can go out on their own. 4. Every Monday, what worked and what changes.: The team reviews the week: replies, visits, Search Console and GA4. It leads with what is working and what to fix, and says what it will do differently next week. 5. Every agent, on every plan.: Seven days free. Paste your website and your team is ready in about a minute, and real people help you set it up. ## Pricing - Solo: $29 a month (or $290 a year), 3,500 credits a month, 1 brand, 1 seat - Growth: $79 a month (or $790 a year), 10,000 credits a month, 3 brands, 3 seats - Custom (talk to us): any amount a month at about 160 credits a dollar ($200 a month is about 32,000 credits), with the brands and seats you need - Free: 250 credits to start, once, 1 brand - Every paid plan has every agent and approvals on WhatsApp; plans differ only in credits, brands and seats - Credits per job: a text post 5, a post with an AI image 30, a reply 2, an article with its image 55, an 8-second clip with sound 704 - Top-ups: 2,000 credits for $10, 9,000 credits for $40, 25,000 credits for $100 - 7-day free trial of Growth with 1,000 credits, no card needed - Details: https://odeist.com/pricing ## Frequently asked questions ### Is there a free trial? Yes: 7 days of Growth with 1,000 credits. You don’t need a card to start. ### How much does Odeist AI cost? Solo is $29 a month for 3,500 credits and Growth is $79 for 10,000. Paying yearly gives you 2 months free. Agencies and bigger teams get a Custom plan at any amount, about 160 credits for every dollar ($200 a month for about 32,000 credits): talk to us. There’s a free plan too, with 250 credits to start. ### What is a credit? One unit for everything an agent does. Each job has a price in credits, worked out from what it costs us to run: a text post is 5, a reply 2, an article with its image 55. Every price is on the pricing page. ### What is the difference between the plans? Every paid plan has every agent. Plans differ only in credits, brands and seats: Solo is 1 brand and 1 seat, Growth 3 brands and 3 seats; a Custom plan has the brands and seats you need. ### What happens when my credits run out? Your agents pause and pick up again when credits come back or you top up. Nothing fails, and a job is never charged by halves. Top-ups start at 2,000 credits for $10, and top-up credits never expire. ### Do approvals on WhatsApp cost extra? No. Every paid plan includes them. Odeist AI sends new posts and replies to your WhatsApp before they go out. Open the link to approve, edit or skip each one. Nothing is published without your say-so. ### I pay per account today. What changes? Nothing you need to do. You keep your plan as it is, and get 150 credits a month for each dollar you pay. ### Can I change or cancel my plan? Yes, from Settings › Plan. A bigger plan starts as soon as its payment goes through, with extra credits for the rest of this month; a smaller one starts when the cycle ends. Cancelling stops the plan at the end of the cycle, and the credits you have stay until they expire. ### Which social networks do you support? Broadcast works on X (Twitter), LinkedIn, Instagram and Facebook. Engage works on X and LinkedIn. Reddit works through Odeist Scout, our app on Reddit, which you install from its app page on Reddit: it finds threads and drafts replies, and you post them yourself. TikTok is coming soon. ### Can Odeist AI find customers on Reddit? Yes, through Odeist Scout, our app on Reddit. Install it in your own subreddit and it finds Reddit threads where people ask for what you sell, reads each subreddit’s rules, drafts a helpful reply you post yourself, and sends a weekly report on what your market says. It’s part of your plan and uses no credits. Reddit reviews every app like it before anyone can install it; Scout is in that review now and usually takes about a week. See odeist.com/docs/reddit. ## What customers say - DataGait: “Odeist AI became our outbound team on X and LinkedIn. It finds data engineers asking the questions DataGait answers and replies in our voice. The Odeist AI team tuned that voice with us on day one.” - The Applier: “Odeist AI puts The Applier in the LinkedIn and X threads where job seekers ask for help. More new users now find us through those replies, and the Odeist AI team is always quick to help.” - the wrkr: “We’re pre-launch and Odeist AI is already growing our waitlist, joining conversations about flaky end-to-end tests. Setup took an afternoon, and the Odeist AI team helped sharpen our message.” - Nuack: “Odeist AI’s scheduled posts keep Nuack’s Instagram, X and LinkedIn alive, and we approve everything from WhatsApp in seconds. It’s like having a social media manager on the team.” - Chivvy: “When Chivvy was two people doing everything, Odeist AI was our social media team. It found people asking for what we’d built, and the Odeist AI team treated our launch like their own.” ## Odeist AI vs Okara Source: https://odeist.com/compare/odeist-vs-okara (Okara details checked on 2026-10-24) Okara gives you a plan. Odeist AI gives you customers, from $29 a month instead of $129. - The job: Odeist AI: Find buyers in live conversations and reply in your voice. Okara: Write a marketing strategy, then draft content across many channels. - Finds people asking for what you sell: Odeist AI: Yes, on X and LinkedIn, on the schedule you set. Okara: Its Reddit agent scans subreddits you pick (Pro plan). - Replies for you: Odeist AI: In your brand voice, posted the moment you approve, or fully on autopilot. Okara: Reddit reply drafts that you post yourself. - Posts for you: Odeist AI: X, LinkedIn, Instagram and Facebook Pages, with AI images and Reels. Okara: X; LinkedIn only on the $249 Pro plan; Instagram marked “soon”. - Approve from: Odeist AI: WhatsApp, in one tap. Okara: A daily feed inside its dashboard. - Starting price: Odeist AI: $29 a month, every agent included. Okara: $129 a month (Lite), $249 for the full suite. - Credits: Odeist AI: 10,000 a month on Growth ($79). Unused plan credits roll over for 1 month. Okara: 2,000 a month on its paid plans, and unused ones expire. 5 per X post, 8 per Reddit thread and reply. - Free trial: Odeist AI: 7 days of Growth with 1,000 credits, no card. Okara: Free plan covers analysis and strategy only. ### Is Odeist AI a good Okara alternative? If you want customers, it’s the better one. Okara spreads itself across strategy, SEO, articles and influencer campaigns. Odeist AI does the job that pays: it finds people asking for what you sell on X and LinkedIn, replies in your voice and keeps your feed busy on X, LinkedIn, Instagram and Facebook Pages. From $29 a month, with 7 days of Growth free. ### Is Odeist AI cheaper than Okara? For most brands, yes. Okara’s paid plans are $129 and $249 a month, each with 2,000 credits. Odeist AI starts at $29 a month, Growth is $79 with 10,000 credits a month, and there’s no card to start. ### Does Odeist AI do Reddit like Okara? Reddit works through Odeist Scout, our app on Reddit, which you install from its app page on Reddit. It finds Reddit threads asking for what you sell and drafts replies you post yourself, as part of your plan. Today, Okara’s Reddit agent sits on its $249 plan and also leaves the posting to you. ## Odeist AI vs Runable Source: https://odeist.com/compare/odeist-vs-runable (Runable details checked on 2026-10-24) Runable can build you a deck. Odeist AI gets you the meeting. - The job: Odeist AI: Find buyers in live conversations and reply for you. Runable: A general AI agent for sites, apps, slides, documents and media. - Finds people asking for what you sell: Odeist AI: Yes, on X and LinkedIn, on the schedule you set, with the reply already written. Runable: Social listening reports on Reddit, X and forums. - Replies in your voice: Odeist AI: Every reply, written as your brand. Runable: Comment replies on its connected social accounts. - Posts for you: Odeist AI: X, LinkedIn, Instagram and Facebook Pages, with AI images and Reels. Runable: Scheduled posts on its connected networks. - How it runs: Odeist AI: Set it once. It works while you sleep. Runable: Starts from your prompt; approve its plan, then it builds. - Approve from: Odeist AI: WhatsApp, in one tap. Runable: Inside the Runable app. - Pricing: Odeist AI: Credit plans from $29 a month, with the price of every job published. Runable: Credit plans at $20 or $100 a month; what each job costs isn’t published. - Free trial: Odeist AI: 7 days of Growth, no card. Runable: A free plan; its limits aren’t on the pricing page. ### Is Odeist AI a good Runable alternative? For growing on social, it’s the sharper tool. Runable is built to do a bit of everything. Odeist AI is built to do one thing brilliantly: find people asking for what you sell on X and LinkedIn, reply in your voice, and keep your feed busy on X, LinkedIn, Instagram and Facebook Pages. ### How does Odeist AI pricing compare with Runable? Both sell credits. Runable’s plans are $20 or $100 a month, and it doesn’t publish what each job costs. Odeist AI’s plans start at $29 a month, and every job’s price is on the pricing page, so you can see what a month will cost before it starts. Try Growth free for 7 days, no card. ### Can Runable find customers on X and LinkedIn? Runable lists social listening reports on Reddit, X and forums. Odeist AI goes further: it finds strangers asking for what you sell on X and LinkedIn and writes the reply that starts the conversation, ready for your one-tap approval. ## Odeist AI vs ReddGrow Source: https://odeist.com/compare/odeist-vs-reddgrow (ReddGrow details checked on 2026-10-18) ReddGrow gives you one network. Odeist AI gives you four, from $29 a month. - Networks: Odeist AI: X, LinkedIn, Instagram and Facebook Pages; Reddit through Odeist Scout. ReddGrow: Reddit. - Finds people asking for what you sell: Odeist AI: Yes, on X and LinkedIn, on the schedule you set. ReddGrow: Scores hot and rising posts in subreddits you choose. - Posts for you: Odeist AI: Posts, AI images and Reels on X, LinkedIn, Instagram and Facebook Pages. ReddGrow: Original Reddit posts that you publish yourself. - Approve from: Odeist AI: WhatsApp, in one tap. ReddGrow: An in-app queue, or Slack. - After you approve: Odeist AI: It goes live, right away. ReddGrow: Queued for its Chrome extension, or copy and post it yourself. - Free trial: Odeist AI: 7 days, no card. ReddGrow: 14 days, card required. ### Is Odeist AI a good ReddGrow alternative? If your buyers live on more than one network, yes. ReddGrow focuses on Reddit and AI search citations. Odeist AI finds buyers on X and LinkedIn, replies in your voice and posts to X, LinkedIn, Instagram and Facebook Pages, and Reddit comes through Odeist Scout, our app on Reddit. ### How much does Odeist AI cost? Plans start at $29 a month, and every paid plan has every agent. The 7-day trial needs no card; ReddGrow asks for a card to start its trial. ### Does Odeist AI work on Reddit? Yes, through Odeist Scout, our app on Reddit, which you install from its app page on Reddit: it finds Reddit threads asking for what you sell and drafts replies you post yourself. Odeist AI also wins customers where B2B buyers talk every day: X and LinkedIn. ## Odeist AI vs Octolens Source: https://odeist.com/compare/odeist-vs-octolens (Octolens details checked on 2026-11-06) Octolens hears more. Odeist AI answers, posts and works on your SEO, from $29 a month. - The job: Odeist AI: Find people asking for what you sell, draft the reply in your voice, and post it when you approve. Octolens: Find and score mentions across many sources, then alert you or your own tools. [3] - Where it listens: Odeist AI: X and LinkedIn conversations, plus Hacker News. Octolens: X, Reddit, Hacker News, LinkedIn, Instagram, YouTube, Bluesky, TikTok, GitHub, Stack Overflow, DEV and review sites; podcasts, news, the web and newsletters on Scale. [4][3] - Replies: Odeist AI: Drafted in your voice and posted on X or LinkedIn when you approve. Octolens: No automated replies, by design: you reply from your own accounts. [3] - Languages: Odeist AI: Writes in the language you choose for each task. Octolens: X listening covers English posts only, no reposts. [4] - Instagram and Facebook: Odeist AI: Publishes your posts with AI images to Instagram and Facebook Pages. Octolens: Instagram: public captions that name your brand, once a day, 1 to 2 days after posting. Facebook: not today. [4][5] - Your posts, SEO and AI answers: Odeist AI: Writes and publishes posts, audits and fixes your site, and checks what AI assistants such as ChatGPT and Gemini say about you. Octolens: Listening only. [3] - API and MCP: Odeist AI: An MCP server for your own AI tools: read your brands and Today, leave drafts. Octolens: API, webhooks and an MCP server on every plan. [1][7] - Starting price: Odeist AI: $29 a month. Octolens: $199 a month, or $159 a month billed yearly (Pro). [2] - Free trial: Odeist AI: 7 days of Growth, no card. Octolens: 7 days, no card. [1] ### Octolens pricing, as of 6 November 2026 - Pro: $199 a month, or $159 a month billed yearly. 15,000 mentions a month, 10 keywords, hourly refresh. - Scale: $599 a month, or $499 a month billed yearly. 40 keywords, real-time refresh, adds podcasts, news, web and newsletters. - Enterprise: Custom. Volume pricing, on request. Octolens counts every mention it collects against the plan. Trial: 7 days, no card. ### Where Octolens is stronger - Far more places to listen. Octolens' docs list about 16 kinds of sources, and its Scale plan adds podcasts, news from 150k+ outlets and the open web. - Noise control. It scores every mention for relevance, and negative keywords and muted authors keep recurring noise out. - Built for your own stack. API, webhooks and an MCP server come with every plan, with Slack and Attio integrations. ### Is Odeist AI an Octolens alternative? For acting on conversations, yes: Odeist AI finds people on X and LinkedIn asking for what you sell, drafts the reply in your voice and posts it when you approve. For listening across many sources, Octolens covers far more: podcasts, news, GitHub, Stack Overflow, review sites and more. ### Does Odeist AI listen as widely as Octolens? No. Today Odeist AI watches X and LinkedIn for people asking for what you sell, and Hacker News. Octolens' platform docs list about 16 kinds of sources, with podcasts, news, the web and newsletters on its Scale plan. ### How do the prices compare? Octolens Pro is $199 a month, or $159 a month billed yearly, with 15,000 mentions and 10 keywords. Odeist AI starts at $29 a month, and Growth is $79 with 10,000 credits. Octolens' trial is 7 days and Odeist AI's is 7 days, both with no card. Prices as of 6 November 2026. ### Sources 1. Octolens pricing: https://octolens.com/pricing (read 2026-11-06) 2. Octolens llms.txt (plan prices): https://octolens.com/llms.txt (read 2026-11-06) 3. Octolens for founders: https://octolens.com/founders (read 2026-11-06) 4. Octolens platforms docs: https://octolens.com/docs/learn/platforms/overview (read 2026-11-06) 5. Octolens for agencies: https://octolens.com/agencies (read 2026-11-06) 6. Octolens home page: https://octolens.com/ (read 2026-11-06) 7. Octolens integrations: https://octolens.com/integrations (read 2026-11-06) ## Odeist AI vs ClimbX Source: https://odeist.com/compare/odeist-vs-climbx (ClimbX details checked on 2026-11-06) ClimbX goes deep on X. Odeist AI posts on four networks and replies on X and LinkedIn, from $29 a month. - Networks: Odeist AI: X, LinkedIn, Instagram and Facebook Pages. ClimbX: X only. [2] - Daily drafts: Odeist AI: Posts on the schedule you set, in your brand's voice, with AI images. ClimbX: 10 Ready Posts a day in your voice, learned from your last 100 to 200 posts. [3][2] - Replies: Odeist AI: Finds people asking for what you sell, drafts the reply and posts it when you approve. ClimbX: Suggests a reply; you edit it and send it on X yourself. [3] - Posting: Odeist AI: Publishes when you approve, or on autopilot. ClimbX: Publishes and schedules the posts you choose; it says it is not an AI auto-poster. [3][2] - Learning: Odeist AI: Learns from what you approve, edit and turn down. ClimbX: Do-more and do-less rules from your own posts' results. [4][6] - Draft score: Odeist AI: Not yet. ClimbX: A score and one clear fix on every draft. [5] - Starting price: Odeist AI: $29 a month. ClimbX: $29 a month (Base: 1 X account, 300 AI credits a day). [1] - Free trial: Odeist AI: 7 days of Growth, no card. ClimbX: 7 days for $1, card on file. [1] ### ClimbX pricing, as of 6 November 2026 - Base: $29 a month. 1 X account, 300 AI credits a day. - Pro: $29 a month for the first 100 customers, then $39. 3 X accounts, 800 AI credits a day. - Agency: $199 a month. 10 X accounts, 8,000 AI credits a day. No free plan: the trial is 7 days for $1 with a card on file. ### Where ClimbX is stronger - It learns from your results. ClimbX turns how your own posts did into do-more and do-less rules for the next drafts. Odeist AI learns from your approvals and edits, not yet from post results. - Every draft gets coached. A score and one clear fix come with each draft. - Depth on X. Its voice comes from your last 100 to 200 posts, and its tools are all about growing an X account. ### Is Odeist AI a good ClimbX alternative? If you market on more than X, yes: Odeist AI posts on X, LinkedIn, Instagram and Facebook Pages, and replies on X and LinkedIn. If X is your only channel and you want each draft scored and coached from your own results, ClimbX goes deeper there. ### Does ClimbX post replies for you? No. Its developer docs say the reply is a suggestion that you edit and send on X yourself. Odeist AI posts the reply when you approve it. ### How do the prices compare? ClimbX starts at $29 a month: Base is $29 for one X account; Pro is $29 for its first 100 customers, then $39. Odeist AI's Solo plan is $29 a month and Growth is $79. ClimbX's trial is 7 days for $1 with a card; Odeist AI's is 7 days with no card. Prices as of 6 November 2026. ### Sources 1. ClimbX pricing: https://climbx.so/pricing (read 2026-11-06) 2. ClimbX llms.txt: https://climbx.so/llms.txt (read 2026-11-06) 3. ClimbX developer docs: https://climbx.so/developers/docs (read 2026-11-06) 4. ClimbX agent skill: https://climbx.so/skills/climbx/SKILL.md (read 2026-11-06) 5. ClimbX home page: https://climbx.so/ (read 2026-11-06) 6. About ClimbX: https://climbx.so/about (read 2026-11-06) ## Odeist AI vs Prefer Source: https://odeist.com/compare/odeist-vs-prefer (Prefer details checked on 2026-11-06) Prefer goes deeper on AI search. Odeist AI adds social and applied SEO fixes, from $29 a month. - The job: Odeist AI: An AI marketing team: social posts and replies, SEO and AI answers. Prefer: AI search visibility: how ChatGPT, Gemini, Perplexity and Google's AI talk about your brand. [1] - AI engines tracked: Odeist AI: ChatGPT and Gemini, with Claude and Grok as options. Prefer: ChatGPT, Gemini, Perplexity, Google AI Overviews and AI Mode on every plan; Claude, Copilot and Grok on Enterprise. [2] - Prompts tracked: Odeist AI: Up to 15 buyer questions a week. Prefer: 150 on the $99 plan. [1] - Fixes to your site: Odeist AI: Applied for you through a GitHub pull request or your CMS, once you approve. Prefer: Action Center actions are manual in the current release. [3] - Articles: Odeist AI: Writes articles and publishes them to WordPress, Webflow, Sanity or GitHub. Prefer: 30 AI articles a month; publishes to WordPress, Ghost, Sanity, Astro, Framer and Strapi, or through a GitHub pull request. [1][4] - Social media: Odeist AI: Writes and publishes posts on X, LinkedIn, Instagram and Facebook Pages; replies on X and LinkedIn. Prefer: Three promotional post variants per article. [5] - Starting price: Odeist AI: $29 a month. Prefer: $99 a month, or $79 a month billed yearly. [1] - Free trial: Odeist AI: 7 days of Growth, no card. Prefer: 7 days, no card, up to 30 prompts. [1] ### Prefer pricing, as of 6 November 2026 - Prefer: $99 a month, or $79 a month billed yearly. 150 tracked prompts, 30 AI articles a month, all 5 engines, 10 competitors per project, unlimited team members. - Prefer Agency: $199 a month, or $179 a month billed yearly. 250 tracked prompts shared across clients, 50 articles a month. No free-forever plan; the trial is 7 days with no card, capped at 30 prompts. Enterprise and a managed service are priced separately. ### Where Prefer is stronger - Deeper AI-search tracking. Five engines on every plan, Perplexity and Google's AI Overviews and AI Mode included, and 150 prompts on the $99 plan. - More places to publish. WordPress, Ghost, Sanity, Astro, Framer and Strapi, plus a GitHub app. - Made for agencies. A pooled-prompt agency plan, with white-label reports on Agency and Enterprise. ### Is Odeist AI a good Prefer alternative? If you want one tool for social, SEO and AI answers, yes. If AI search visibility is the whole job, Prefer tracks more: five engines including Perplexity and Google's AI surfaces, and 150 prompts on its $99 plan. ### Does Odeist AI track Perplexity or Google AI Overviews? Not today. Odeist AI asks ChatGPT and Gemini by default, with Claude and Grok as options, up to 15 buyer questions a week. Prefer tracks ChatGPT, Gemini, Perplexity, Google AI Overviews and AI Mode on every plan. ### How do the prices compare? Prefer is $99 a month, or $79 a month billed yearly; Prefer Agency is $199, or $179 yearly. Odeist AI starts at $29 a month, and Growth is $79. Prefer's trial is 7 days and Odeist AI's is 7 days, both with no card. Prices as of 6 November 2026. ### Sources 1. Prefer facts (llm-info): https://tryprefer.com/llm-info/ (read 2026-11-06) 2. Prefer docs: engines: https://tryprefer.com/docs/concepts/engines (read 2026-11-06) 3. Prefer docs: Action Center: https://tryprefer.com/docs/act/action-center (read 2026-11-06) 4. Prefer docs: publish to your CMS: https://tryprefer.com/docs/act/publish-to-your-cms (read 2026-11-06) 5. Prefer docs: Content Studio: https://tryprefer.com/docs/act/content-studio (read 2026-11-06) 6. Prefer agency pricing: https://tryprefer.com/pricing/agencies/ (read 2026-11-06) ## Odeist AI vs Explee Source: https://odeist.com/compare/odeist-vs-explee (Explee details checked on 2026-11-06) Explee writes to strangers. Odeist AI answers people already asking, from $29 a month. - The job: Odeist AI: Find people already asking for what you sell, reply and post for you. Explee: Find companies and people that match your criteria and run email campaigns to them. [1] - Who it reaches: Odeist AI: People talking in public on X and LinkedIn. Explee: Companies and people from its database, searched in plain language. [1] - Writing: Odeist AI: Replies and posts in your brand's voice. Explee: Writes each first email from your saved instructions, or sends your own copy. [1] - Sending: Odeist AI: From your own X and LinkedIn accounts, when you approve. Explee: From Explee's shared mailboxes, or mailboxes bought for your own domain. [1] - Approval: Odeist AI: Every reply and post waits for your yes, unless you choose autopilot. Explee: Lead lists go to review; with Auto-manage campaigns on, they are approved automatically. [1] - Contact data: Odeist AI: None. Explee: Finds work emails and phone numbers. [1] - Pricing unit: Odeist AI: Credit plans from $29 a month. Explee: Credits at $0.01 each, spent per company, person, email or phone found. [1] ### Explee pricing, as of 6 November 2026 - Credits: 1 credit = $0.01. Company search: 0.5 credits a company after the first 10. People search: 1 credit a person after the first 10. - Contact details: Per result found. Email: 1.5 credits (basic) or 5 (premium). Phone: 15 credits (basic) or 30 (premium). From Explee's public API reference. Its pricing page didn't load for us on 6 November 2026, so monthly plan prices aren't shown here; check explee.com/pricing. ### Where Explee is stronger - Outbound at scale. It searches companies and people in plain language and runs email campaigns to them. - Contact data. It finds work emails and phone numbers, which Odeist AI doesn't. - Hands-off campaigns. With Auto-manage campaigns on, lead lists are approved for you. ### Is Odeist AI an Explee alternative? Only if you want warm conversations instead of cold email. Odeist AI doesn't send cold email or find contact details. It finds people on X and LinkedIn already asking for what you sell and replies in your voice when you approve. ### Where do Explee's emails come from? Its API reference says emails go out from Explee's shared mailboxes, or from mailboxes bought for your own domain. ### How does Explee charge? In credits worth $0.01 each, per company, person, email or phone found, per its API reference. Odeist AI's plans start at $29 a month, with a 7-day trial and no card. As of 6 November 2026. ### Sources 1. Explee public API reference (OpenAPI): https://api.explee.com/public/api/openapi.json (read 2026-11-06) ## Odeist AI vs Relinta Source: https://odeist.com/compare/odeist-vs-relinta (Relinta details checked on 2026-11-06) Relinta runs LinkedIn outreach. Odeist AI answers buyers who ask on X and LinkedIn and posts on four networks, from $29 a month. - The job: Odeist AI: Find people asking for what you sell, reply in public and post for you. Relinta: Automate LinkedIn outreach, follow-ups, replies and reporting across sender accounts. [1] - Networks: Odeist AI: X, LinkedIn, Instagram and Facebook Pages. Relinta: LinkedIn. [1] - Outreach: Odeist AI: Public replies to posts that ask for what you sell; no invites or DMs. Relinta: Invite, then up to three follow-ups, from playbooks. [1] - Comments: Odeist AI: Reply drafts on conversations it finds, approved one by one. Relinta: Auto-comments on accounts you pick; with Review first on, you approve, edit or skip each. [1][2] - Answer from your chat app: Odeist AI: Approve in WhatsApp, Telegram or Slack. Relinta: Answer replies from Telegram or Slack. [2] - API and MCP: Odeist AI: An MCP server for your own AI tools. Relinta: API and MCP on every plan, Starter included. [1][2] - Starting price: Odeist AI: $29 a month. Relinta: Free for 1 LinkedIn account and 1,000 leads; Pro is $79 per LinkedIn account a month. [1] ### Relinta pricing, as of 6 November 2026 - Starter: $0. 1 LinkedIn account, up to 1,000 leads, API and MCP access. - Pro: $79 per LinkedIn account a month. Priced per LinkedIn sender account. - Team: $299 per 10 LinkedIn accounts a month. $29.90 per account. ### Where Relinta is stronger - LinkedIn outreach across many senders, with duplicates blocked and warm replies shown first. - A free plan. Starter is $0 for one LinkedIn account and up to 1,000 leads. - Low cost per seat for teams: $29.90 a LinkedIn account on Team. ### Is Odeist AI a Relinta alternative? For being found by buyers who are already asking, yes. Odeist AI doesn't send LinkedIn invites or DMs; it replies in public to posts asking for what you sell, on X and LinkedIn, when you approve. For multi-sender LinkedIn outreach, Relinta is built for it. ### Does Odeist AI send LinkedIn connection requests or DMs? No. It writes public replies and posts, and each one waits for your approval unless you turn on autopilot. ### How do the prices compare? Relinta's Starter plan is free for one LinkedIn account and 1,000 leads; Pro is $79 per LinkedIn account a month; Team is $299 per 10 accounts. Odeist AI starts at $29 a month for four networks, with a 7-day trial and no card. Prices as of 6 November 2026. ### Sources 1. Relinta home page and pricing: https://relinta.com/ (read 2026-11-06) 2. Relinta changelog: https://relinta.com/changelog (read 2026-11-06) ## Odeist AI vs Postiz Source: https://odeist.com/compare/odeist-vs-postiz (Postiz details checked on 2026-11-06) Postiz reaches more channels. Odeist AI writes, replies and can wait for your approval, from $29 a month. - The job: Odeist AI: Write and publish your posts, find conversations and reply, work on your SEO. Postiz: Schedule and publish posts to 30+ channels. [5] - Networks: Odeist AI: X, LinkedIn, Instagram and Facebook Pages. Postiz: 30+ channels. [5] - Approval: Odeist AI: A post goes out only after your yes, unless you choose autopilot. Postiz: A public preview page per post for comments; its docs say approval is a conversation, not a state. [6] - Replies: Odeist AI: Finds people asking for what you sell and drafts replies on X and LinkedIn. Postiz: Its agent can't read or reply to comments on your posts. [7] - Open source: Odeist AI: Hosted for you. Postiz: AGPL-3.0; you can host it yourself. [4] - MCP: Odeist AI: An MCP server for your own AI tools. Postiz: An MCP server with 16 tools. [8] - Starting price: Odeist AI: $29 a month. Postiz: $29 a month for 5 channels, or $23 a month billed yearly. [1] - Free trial: Odeist AI: 7 days of Growth, no card. Postiz: 7 days, with a payment method; no free plan. [2][3] ### Postiz pricing, as of 6 November 2026 - Standard: $29 a month, or $23 billed yearly. 5 channels. - Team: $39 a month, or $31 billed yearly. 10 channels. - Pro: $49 a month, or $39 billed yearly. 30 channels. - Ultimate: $99 a month, or $79 billed yearly. 100 channels. The hosted trial needs a payment method, and there is no free hosted plan. The open-source edition can be self-hosted. ### Where Postiz is stronger - Many more channels. Postiz publishes to 30+ social channels, from Bluesky and Threads to YouTube and Pinterest. - Open source. It is AGPL-3.0 with about 41k GitHub stars, and you can host it yourself. - Video clips. Its agent turns a long YouTube video into captioned vertical clips. ### Is Odeist AI a good Postiz alternative? If you want the writing and replying done for you, yes. If you need 30+ channels or self-hosting, Postiz is built for that, and Odeist AI covers X, LinkedIn, Instagram and Facebook Pages. ### Does Postiz have an approval step? Postiz gives every post a public preview page where people can comment. Its approvals doc calls approval a conversation, not a state. In Odeist AI, a post waits for your yes before it goes out, unless you choose autopilot. ### How do the prices compare? Postiz Standard is $29 a month for 5 channels ($23 billed yearly), up to Ultimate at $99 for 100 channels. Odeist AI starts at $29 a month and writes the posts too. Postiz's hosted trial needs a payment method; Odeist AI's doesn't. Prices as of 6 November 2026. ### Sources 1. Postiz pricing: https://postiz.com/pricing (read 2026-11-06) 2. Postiz docs: trial: https://docs.postiz.com/cloud/trial (read 2026-11-06) 3. Postiz docs: plans: https://docs.postiz.com/cloud/plans (read 2026-11-06) 4. Postiz on GitHub: https://github.com/gitroomhq/postiz-app (read 2026-11-06) 5. Postiz home page: https://postiz.com/ (read 2026-11-06) 6. Postiz docs: approvals: https://docs.postiz.com/general/approvals (read 2026-11-06) 7. Postiz docs: agents: https://docs.postiz.com/general/agents (read 2026-11-06) 8. Postiz docs: MCP tools: https://docs.postiz.com/mcp/tools (read 2026-11-06) ## Odeist AI vs 1eye Source: https://odeist.com/compare/odeist-vs-1eye (1eye details checked on 2026-11-06) 1eye tells you who visited. Odeist AI answers who's asking, from $29 a month. - The job: Odeist AI: Find people asking for what you sell in public, reply and post for you. 1eye: Turn anonymous website visits, in-market activity and LinkedIn engagement into named buyers. [1] - Signals: Odeist AI: Public posts on X and LinkedIn, and Hacker News. 1eye: Website visitors, in-market activity, LinkedIn engagement. [1] - Fit: Odeist AI: Ranks each conversation against your brand documents before drafting. 1eye: Matches buyers to your ideal customer and shows the reasoning. [1][5] - What happens next: Odeist AI: A reply drafted in your voice, waiting for your yes. 1eye: Buyers go to HubSpot, Salesforce, Slack or a webhook. [3] - API and MCP: Odeist AI: An MCP server for your own AI tools. 1eye: An MCP server with 30 tools. [4] - Starting price: Odeist AI: $29 a month. 1eye: Free with 500 credits at sign-up; then $0.05 a credit, 10 credits a buyer. [1][2] ### 1eye pricing, as of 6 November 2026 - Free: $0. 500 credits at sign-up, up to 5 users. - Prepaid: $0.05 a credit. 10 credits a qualified buyer, 1 credit a buyer signal. - High Volume: Custom. Volume tiers and custom invoicing. 1eye's credits docs say 10 credits a qualified buyer in the text and 5 in a table; its home page says 10. ### Where 1eye is stronger - It sees your website. 1eye names buyers behind anonymous visits, which Odeist AI doesn't. - CRM ready. HubSpot, Salesforce, Slack and webhooks are live destinations. - Explained fit. Each match comes with the reasoning against your ideal customer. ### Is Odeist AI a 1eye alternative? Only for a different kind of signal. 1eye identifies buyers behind website visits; Odeist AI doesn't. Odeist AI finds people asking in public on X and LinkedIn and replies for you when you approve. ### Does Odeist AI identify website visitors? No. Odeist AI works from public conversations on X and LinkedIn, and Hacker News. ### How do the prices compare? 1eye starts free with 500 credits, then $0.05 a credit, 10 credits a qualified buyer on its home page. Odeist AI starts at $29 a month, with a 7-day trial and no card. Prices as of 6 November 2026. ### Sources 1. 1eye home page: https://www.1eye.ai/ (read 2026-11-06) 2. 1eye docs: credits: https://docs.1eye.ai/get-started/credits-usage (read 2026-11-06) 3. 1eye destinations: https://www.1eye.ai/destinations (read 2026-11-06) 4. 1eye docs: MCP: https://docs.1eye.ai/mcp/introduction (read 2026-11-06) 5. 1eye docs: match: https://docs.1eye.ai/buyer-intelligence/match (read 2026-11-06) ## How do I get my business into ChatGPT and other AI answers? Source: https://odeist.com/learn/how-to-get-your-business-into-ai-answers (updated 2026-11-06) Make sure AI search crawlers can read your site, publish pages that answer your buyers’ questions plainly with facts and sources, get your business named on the pages assistants already cite, and check every week which brands the assistants name. There is no special file or markup that switches it on. ### 1. Let the search crawlers in Assistants that search the web use their own bots. OpenAI uses OAI-SearchBot to surface websites in ChatGPT search, separately from GPTBot, which collects training data; blocking GPTBot doesn’t remove you from search, because each setting is independent. Anthropic’s Claude-SearchBot indexes pages for Claude’s search, and blocking it may reduce your visibility in its answers. Perplexity’s PerplexityBot surfaces and links websites in Perplexity’s results. For Google’s AI Overviews and AI Mode, a page has to be indexed and eligible to show in Search with a snippet; Google says there are no extra technical requirements. Check your robots.txt and make sure none of these are blocked by an old rule. ### 2. Answer the questions your buyers ask Write down the ten questions a buyer asks before choosing a product like yours, in their words. Give each one a page or a section that answers it in the first paragraph, then backs it up. Google’s advice for its AI features is to focus on unique, non-commodity content that people find helpful. Research on generative engines found that adding citations, quotations and statistics raised a page’s visibility in AI answers the most, while keyword stuffing gave little or no improvement. ### 3. Get named where assistants already look Ask an assistant your buyers’ questions and look at the pages it cites: comparison posts, “best tools” lists, review sites, forums. If those pages name your competitors and not you, that is your outreach list. Ask the author to consider your product, with a short, honest reason it belongs there. ### 4. Measure it every week Ask the same questions on the same assistants every week and count who gets named. Judge the trend over several weeks, not a single answer. ### How Odeist AI helps Odeist AI’s AI answers agent does steps 1 and 4 for you: it asks ChatGPT, Gemini, Claude or Grok your buyers’ questions every week, shows how often each brand is named, checks whether AI crawlers may read your site, and turns what it finds into fixes and article ideas in Today. Its Articles agent can write the pages for step 2. ### Do I need an llms.txt file to appear in AI answers? No. Google says you don’t need new machine-readable files, AI text files or markup to appear in its AI Overviews or AI Mode. An llms.txt file is a proposal that some AI tools read; it is cheap to add, but it is not required. ### Will blocking GPTBot keep me out of ChatGPT? Not out of ChatGPT search. OpenAI says GPTBot is for training data and OAI-SearchBot is for search results, and each setting is independent. Block OAI-SearchBot and you leave ChatGPT’s search answers. ### How long does it take? Crawler changes can be quick: OpenAI says its search systems can take about 24 hours to adjust after a robots.txt change. New pages and mentions on other sites take longer to show up in answers, so measure over weeks. ### Sources 1. OpenAI: Overview of OpenAI crawlers: https://developers.openai.com/api/docs/bots (read 2026-11-06) 2. Anthropic: Does Anthropic crawl data from the web, and how can site owners block the crawler?: https://support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler (read 2026-11-06) 3. Perplexity: Perplexity crawlers: https://docs.perplexity.ai/guides/bots (read 2026-11-06) 4. Google Search Central: AI features and your website: https://developers.google.com/search/docs/appearance/ai-features (read 2026-11-06) 5. Google Search Central blog: Top ways to ensure your content performs well in Google’s AI experiences on Search (21 May 2025): https://developers.google.com/search/blog/2025/05/succeeding-in-ai-search (read 2026-11-06) 6. Aggarwal et al., “GEO: Generative Engine Optimization”, full text, section 4: https://arxiv.org/html/2311.09735v3 (read 2026-11-06) ## Do I need an llms.txt file? Source: https://odeist.com/learn/do-i-need-an-llms-txt-file (updated 2026-11-06) You don’t need one, but it is cheap to add. llms.txt is a proposal for a Markdown file that gives AI tools a short summary of your site and links to its best pages. Google says you don’t need AI text files to appear in its AI Overviews or AI Mode, so treat llms.txt as a small extra for AI tools that read it, not a ranking switch. ### What the file is Jeremy Howard proposed llms.txt on 3 September 2024. It is a Markdown file, usually at /llms.txt, and the only required part is an H1 with the name of the site or project. After that come a blockquote with a short summary, any notes you want, and H2 sections that list links with a name, a URL and an optional note. ### What it doesn’t do Search engines have not adopted it as a standard; the proposal itself presents it as a curated overview for language models, different from a sitemap. Google says you don’t need new machine-readable files, AI text files or markup to appear in its AI features. It won’t replace pages that answer your buyers’ questions. ### How to write one in ten minutes 1. Start with “# Your company name”. 2. Add one line in a blockquote that says what you sell and who for, in plain words. 3. Add “## Product”, “## Pricing” and “## Guides” sections, each a list of links to your real pages with a few words on each. 4. Save it as llms.txt at the root of your site and check it opens at yoursite.com/llms.txt. 5. Update it when your prices or main pages change. ### How Odeist AI helps Odeist AI’s AI answers agent checks whether your site has an llms.txt file with a title, a summary and links, as part of its check that AI tools can read your site. ### Is llms.txt the same as robots.txt? No. robots.txt tells crawlers which paths they may fetch. llms.txt is a summary and reading list for AI tools; it doesn’t allow or block anything. ### Does odeist.com have one? Yes, at odeist.com/llms.txt, with a longer version at odeist.com/llms-full.txt. ### Sources 1. llmstxt.org: The /llms.txt file: https://llmstxt.org/ (read 2026-11-06) 2. Google Search Central: AI features and your website: https://developers.google.com/search/docs/appearance/ai-features (read 2026-11-06) ## Should I block AI crawlers in robots.txt? Source: https://odeist.com/learn/should-you-block-ai-crawlers (updated 2026-11-06) Most businesses that want customers should allow the AI search bots and decide separately about the training bots. Search bots, such as OpenAI’s OAI-SearchBot, Anthropic’s Claude-SearchBot and PerplexityBot, decide whether you can be found and linked in AI answers. Training bots, such as GPTBot and ClaudeBot, collect content for model training. You can block one kind and allow the other. ### Search bots: what blocking costs you - OpenAI’s OAI-SearchBot is used to surface websites in ChatGPT’s search features; sites that opt out don’t appear in ChatGPT search answers, though they can still show as navigational links. - Anthropic’s Claude-SearchBot improves search results for Claude’s users; blocking it means your content isn’t indexed for search, which may reduce your visibility in Claude’s answers. - PerplexityBot surfaces and links websites in Perplexity’s search results and isn’t used to train foundation models. ### Training bots: your call - OpenAI’s GPTBot crawls content that may be used to train its models. OpenAI says each setting is independent, so you can block GPTBot and still allow OAI-SearchBot. - Anthropic’s ClaudeBot collects content that could contribute to training; blocking it excludes your future content from training data. - Google-Extended is a robots.txt token, not a separate crawler. It controls use of your content for Gemini models and, Google says, doesn’t affect your inclusion or ranking in Google Search. ### Fetchers that act for a person Some fetchers visit a page because a person asked about it. Perplexity says Perplexity-User generally ignores robots.txt for that reason. OpenAI says ChatGPT-User isn’t used to decide what appears in search, and robots.txt rules may not apply to it. robots.txt is a request, not a lock: the protocol itself says its rules are not a form of access authorization. ### A robots.txt that allows search and blocks training Each group names a bot with User-agent and lists what it may not fetch with Disallow. A crawler follows the group that names it, and the “*” group only when no group does. So to block GPTBot, ClaudeBot and Google-Extended while allowing everything else, give them a group of their own with “Disallow: /” and leave the search bots under your normal rules. A bot named in its own group ignores the “*” rules, so repeat any Disallow lines it should still follow. Test the file after every change. ### How Odeist AI helps Odeist AI’s AI answers agent reads your robots.txt and tells you if it blocks the crawlers behind the assistants it checks, so an old rule doesn’t quietly keep you out of AI answers. ### How quickly does ChatGPT notice a robots.txt change? OpenAI says it can take about 24 hours from a robots.txt update for its search systems to adjust. ### Does Anthropic follow robots.txt? Anthropic says its bots honor robots.txt directives, and notes that IP blocking may not work reliably because it stops the bots reading the file at all. ### Sources 1. OpenAI: Overview of OpenAI crawlers: https://developers.openai.com/api/docs/bots (read 2026-11-06) 2. Anthropic: Does Anthropic crawl data from the web, and how can site owners block the crawler?: https://support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler (read 2026-11-06) 3. Perplexity: Perplexity crawlers: https://docs.perplexity.ai/guides/bots (read 2026-11-06) 4. Google Search Central: Google’s common crawlers (Google-Extended): https://developers.google.com/search/docs/crawling-indexing/google-common-crawlers (read 2026-11-06) 5. RFC 9309: Robots Exclusion Protocol: https://www.rfc-editor.org/rfc/rfc9309 (read 2026-11-06) ## How do I find customers on X and LinkedIn without spamming people? Source: https://odeist.com/learn/find-customers-on-x-and-linkedin-without-spamming (updated 2026-11-06) Search for posts where people ask for what you sell, reply only where you can genuinely help, answer their question before you mention your product, and read every reply before it goes out. One useful reply to the right post does more than a stream of pitches, and the platforms are acting against the stream. ### Search for questions, not keywords Search the phrases buyers use when they need something: “recommend a tool for…”, “alternative to…”, “how do you handle…”, plus your category and your competitors’ names. Skip posts that only mention a keyword in passing. Go often. On X, the standard recent search in X’s API covers the last 7 days of posts, so a conversation that is a week old is already hard to find for many tools. Daily is a good rhythm. ### Reply like a person who knows the subject 1. Answer the question in the first line, with something useful even if they never buy from you. 2. Mention your product only if it fits what they asked, and say plainly that it’s yours. 3. Keep it short. No link unless it helps; on X a link takes 23 of your 280 characters. 4. If you can’t add anything, don’t reply. ### Know what the platforms don’t allow LinkedIn’s Professional Community Policies don’t allow untargeted, irrelevant or obviously unwanted commercial or promotional messages, or using invitations to send promotional messages to people you don’t know. In March 2026 LinkedIn said it is working to curb comment automation and make engagement pods ineffective. Reading every reply before it goes out keeps you on the right side of both. ### How Odeist AI helps Odeist AI’s reply agents search X and LinkedIn for posts about your space on the schedule you set, skip the ones that don’t fit your brand, and draft a reply in your voice. With approvals on, each draft waits for you in Today, WhatsApp, Telegram or Slack, where you approve, edit or skip it. Each reply drafted is 2 credits. ### Should I send a direct message instead? Usually not at first. A public reply helps everyone reading the thread and doesn’t land uninvited in someone’s inbox. If they reply and ask, then move to a message. ### How many replies a day is too many? There is no safe number to aim for. The question is whether each reply helps the person you reply to. Ten good ones beat a hundred quick ones. ### Sources 1. X developer platform: Search Posts (introduction): https://docs.x.com/x-api/posts/search/introduction (read 2026-11-06) 2. X developer platform: Counting characters: https://docs.x.com/fundamentals/counting-characters (read 2026-11-06) 3. LinkedIn Professional Community Policies: https://www.linkedin.com/legal/professional-community-policies (read 2026-11-06) 4. LinkedIn: How LinkedIn Is Improving the Feed to Show More Relevant, Authentic Professional Content (12 March 2026): https://news.linkedin.com/2026/ImprovingTheFeed (read 2026-11-06) ## How do I write a brand voice guide? Source: https://odeist.com/learn/how-to-write-a-brand-voice-guide (updated 2026-11-06) Fit it on one page: place your brand on four tone dimensions, list three words that describe how you sound and three that don’t, list words you always and never use, and show two or three before-and-after examples. A short guide that people, and AI tools, actually follow beats a long one nobody reads. ### Voice stays, tone moves Mailchimp’s content style guide puts it simply: you have the same voice all the time, but your tone changes. Your guide should describe the voice and say how the tone shifts, for example in a launch post versus a reply to a complaint. ### The one-page guide 1. Place your brand on Nielsen Norman Group’s four dimensions of tone: formal or casual, serious or funny, respectful or irreverent, matter-of-fact or enthusiastic. 2. Write three words you are (for example: direct, warm, specific) and three you are not (salesy, vague, smug). 3. List words you always use (your product’s real names, your customers’ words for the problem) and words you never use (jargon, hype, competitors’ names in posts). 4. Write two or three before-and-after examples: a bland sentence, then the same sentence in your voice. 5. Add how the tone changes for three moments: a launch, a reply to a question, a reply to a complaint. ### Keep it alive Every time you rewrite a draft, ask why. If the reason is general (“we never say ‘leverage’”), add it to the guide. Within a month the guide sounds like you. ### How Odeist AI helps Odeist AI writes a Voice document for each brand from its website, and every post and reply follows it. You can edit it like any document, rate sample posts in Tune your voice, and your skips and edits become learned rules for each agent that you can read and change. ### How long should a voice guide be? One page is enough for most small teams. If it doesn’t fit, it is probably a style manual, which is useful too, but separate. ### Can an AI tool follow it? Yes, if it is concrete. Lists of words to use and avoid, and before-and-after examples, work far better than adjectives alone. ### Sources 1. Mailchimp Content Style Guide: Voice and tone: https://styleguide.mailchimp.com/voice-and-tone/ (read 2026-11-06) 2. Nielsen Norman Group: The Four Dimensions of Tone of Voice (Kate Moran): https://www.nngroup.com/articles/tone-of-voice-dimensions/ (read 2026-11-06) ## Should I approve every AI-written post before it goes out? Source: https://odeist.com/learn/should-you-approve-every-ai-post (updated 2026-11-06) At first, yes. Approve every draft until the edits stop, then let the steady, low-risk kinds of posts run on their own and keep approving replies and comments. A reply goes to a real person in a real conversation, and platforms such as LinkedIn act against automated engagement. ### Why approve - Facts: an AI can state a number or a feature you don’t have. You catch it in seconds; a customer catches it in public. - Fit: a reply to a post that isn’t really about your product is noise, however well written. - Platform rules: LinkedIn said in March 2026 that it takes action against behaviors that promote inauthentic engagement, including comment automation. - Search: Google says using automation, including AI, to produce content mainly to manipulate rankings violates its spam policies. ### When to let an agent run on its own 1. Approve everything for the first two weeks and note why you edit. 2. Turn your edits into rules: words to avoid, facts to always include, topics to skip. 3. When a week passes with almost no edits, let that one agent post on its own. Start with scheduled posts, not replies. 4. Keep reading what goes out. Turn approvals back on the moment something reads wrong. ### Disclose relationships If teammates post about your product, the FTC’s guidance says they should disclose that they work for you; listing the employer on a profile page isn’t enough. ### How Odeist AI helps Each Odeist AI agent has its own switch for whether it asks you first. Drafts that wait for you show in Today and, when connected, in WhatsApp, Telegram or Slack. Approve, edit or skip with one tap; a skip with a reason teaches the agent, and your edits become learned rules you can read and change. ### Isn’t approving everything slow? It takes seconds per draft if the drafts arrive where you already are, such as your phone. The time goes down as the drafts need fewer edits. ### Do I have to say a post was written with AI? Rules differ by platform and country. Google says AI or automation disclosures are useful where readers might ask how content was made. Check each platform’s current rules for ads and sponsored posts. ### Sources 1. LinkedIn: How LinkedIn Is Improving the Feed to Show More Relevant, Authentic Professional Content (12 March 2026): https://news.linkedin.com/2026/ImprovingTheFeed (read 2026-11-06) 2. Google Search Central: Creating helpful, reliable, people-first content: https://developers.google.com/search/docs/fundamentals/creating-helpful-content (read 2026-11-06) 3. FTC: FTC’s Endorsement Guides: What People Are Asking: https://www.ftc.gov/business-guidance/resources/ftcs-endorsement-guides-what-people-are-asking (read 2026-11-06) ## How do credits work in an AI marketing tool, and how many do I need? Source: https://odeist.com/learn/how-do-ai-marketing-credits-work (updated 2026-11-06) Credits are one balance that pays for every job: each post, reply, article or clip costs a set number of credits. To know how many you need, list a normal month (how many posts, replies and articles), multiply each by its price, and add a margin for busy weeks. ### Estimate a month in three steps 1. List what you want each month: for example 12 posts, 60 replies and 4 articles for one brand. 2. Multiply each by its credit price from the tool’s price list. 3. Add about a fifth for busy weeks, then pick the plan whose monthly credits cover it. ### What to check in any credit plan - Is every price published, or only a few examples? - What happens to unused credits at the end of the month? - What happens when you run out: do jobs fail, or pause and wait? - Can you buy more without changing plan, and do those credits expire? ### How Odeist AI helps Odeist AI publishes every price on its pricing page: a text post is 5 credits, a post with an AI image 30, a reply 2 and an article with its image 55. Solo gives 3,500 credits a month for $29 and Growth 10,000 for $79. Unused plan credits roll over 1 month, up to one month’s worth; top-ups start at 2,000 credits for $10 and never expire; when credits run out, agents pause and pick up again when credits come back. You can try Growth free for 7 days with 1,000 credits, no card needed. ### Why credits instead of a price per feature? Jobs cost very different amounts to run. One balance lets you spend more on video one month and more on replies the next, without changing plans. ### Do approvals or scheduling use credits on Odeist AI? No. The morning rundown and digests, your activity, applying a fix, publishing an article to your site, scheduling, connecting accounts, your team and share links are never charged. ### Sources 1. Odeist AI pricing (every price, rollover and top-ups): https://odeist.com/pricing (read 2026-11-06) ## How do I see which posts bring visitors to my website? Source: https://odeist.com/learn/how-to-see-which-posts-bring-visitors (updated 2026-11-06) Tag every link you share with UTM parameters, at least utm_source, utm_medium and utm_campaign, use the same lowercase values every time, and read the results in your analytics by source and campaign. For search, connect Google Search Console. ### Tag your links Google Analytics asks you to always use utm_source, utm_medium and utm_campaign; missing parameters show up as (not set) in reports. A tagged link looks like this: yoursite.com/pricing?utm_source=linkedin&utm_medium=social&utm_campaign=launch-week. ### Keep the names consistent - Lowercase only: linkedin, x, newsletter. - One medium for all social posts, such as social. - A campaign name per push, such as launch-week or pricing-change. ### Add search Search Console is Google’s free service for your site’s presence in Google Search: the searches you appear for, click-through rates and indexing problems. Together with tagged links, it shows both halves: who found you by searching and who came from your posts. ### How Odeist AI helps Odeist AI’s weekly report reads Search Console and GA4 when you connect them, and says what worked and what changes next week. ### Do UTM tags hurt SEO? Use them on links you share in posts, emails and bios, not on links between your own pages, where they would split your own visits into fake sources. ### Sources 1. Google Analytics Help: URL builders: Collect campaign data with custom URLs: https://support.google.com/analytics/answer/10917952 (read 2026-11-06) 2. Search Console Help: About Search Console: https://support.google.com/webmasters/answer/9128668 (read 2026-11-06) ## Get your business into AI answers: a 30-day playbook Source: https://odeist.com/learn/ai-search-visibility-playbook (updated 2026-11-06) In four weeks: measure which brands AI assistants name for your buyers’ questions, make sure their search crawlers can read your site, publish pages that answer those questions with facts and sources, then get your business named on the pages the assistants already cite. Measure again at the end. ### Week 1: measure where you stand 1. Write ten questions a buyer asks before choosing a product like yours, in their words. Include “best … for …” and “alternative to …” questions. 2. Ask each one on ChatGPT, Gemini, Claude and Perplexity. Note which brands each answer names, in what order, and which pages it cites. 3. Count your share: your mentions divided by all brand mentions. This is your baseline. ### Week 2: let the right crawlers in 1. Open yoursite.com/robots.txt. Make sure OAI-SearchBot, Claude-SearchBot and PerplexityBot aren’t blocked: they decide whether you can appear and be linked in those assistants’ search answers. 2. Decide separately about training bots such as GPTBot and ClaudeBot; OpenAI says each of its settings is independent. 3. Verify your site in Google Search Console and fix pages that aren’t indexed. Google’s AI Overviews and AI Mode only link to pages that are indexed and eligible to show with a snippet. 4. Optional: add an llms.txt file with a one-line summary and links to your key pages. ### Week 3: publish answers worth citing 1. For each of your ten questions, write or rewrite one page that answers it in the first paragraph. 2. Back the answer with specifics: your own numbers, a short quote from a customer, and links to sources. In the GEO research, citing sources, adding quotations and adding statistics raised visibility in AI answers the most; keyword stuffing did little. 3. Keep it original. Google’s advice for its AI features is to focus on unique, non-commodity content. Don’t publish dozens of thin AI-written pages: Google treats generating many pages without adding value as scaled content abuse. 4. Keep facts current: prices, plans and features that change should be updated on the same day. ### Week 4: get named where assistants look 1. Take the pages the assistants cited in week 1. Sort them into lists (“best tools for…”), reviews, forums and comparisons. 2. For each list or comparison that names competitors and not you, send the author a short, honest note on why your product belongs there. 3. Ask a few happy customers for a review on the sites that came up. 4. Ask your ten questions again and compare with week 1. Repeat monthly. ### How Odeist AI helps Odeist AI can run most of this: its AI answers agent asks your buyers’ questions on ChatGPT, Gemini, Claude or Grok every week, tracks how often each brand is named, checks your robots.txt, llms.txt and structured data, and hands you fixes and article ideas in Today. The SEO agent audits your site daily, and the Articles agent writes the answer pages in your voice for you to approve. ### Do I need to do this on every assistant? Start with the ones your buyers use. The crawler and page work in weeks 2 and 3 helps on all of them at once. ### When will I see results? Crawler fixes can show within days; OpenAI says its search systems adjust to a robots.txt change in about 24 hours. New pages and mentions take weeks, which is why you measure monthly. ### Sources 1. OpenAI: Overview of OpenAI crawlers: https://developers.openai.com/api/docs/bots (read 2026-11-06) 2. Anthropic: Does Anthropic crawl data from the web, and how can site owners block the crawler?: https://support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler (read 2026-11-06) 3. Perplexity: Perplexity crawlers: https://docs.perplexity.ai/guides/bots (read 2026-11-06) 4. Google Search Central: AI features and your website: https://developers.google.com/search/docs/appearance/ai-features (read 2026-11-06) 5. Search Console Help: About Search Console: https://support.google.com/webmasters/answer/9128668 (read 2026-11-06) 6. llmstxt.org: The /llms.txt file: https://llmstxt.org/ (read 2026-11-06) 7. Aggarwal et al., “GEO: Generative Engine Optimization”, full text, section 4: https://arxiv.org/html/2311.09735v3 (read 2026-11-06) 8. Google Search Central blog: Top ways to ensure your content performs well in Google’s AI experiences on Search (21 May 2025): https://developers.google.com/search/blog/2025/05/succeeding-in-ai-search (read 2026-11-06) 9. Google Search Central: Spam policies for Google web search (scaled content abuse): https://developers.google.com/search/docs/essentials/spam-policies (read 2026-11-06) ## Build in public on X: a 4-week playbook Source: https://odeist.com/learn/build-in-public-on-x-playbook (updated 2026-11-06) Post three times a week about what you built, what you learned and what you are deciding, with real details. Reply to people in your space every day. Link to your product only when it helps, track which posts bring visitors, and launch with a story, not an announcement. ### Week 1: set up 1. Write your profile in one line: what you’re building and for whom. 2. Pin a post that says what you’re building, why, and what you’ll share. 3. List 20 people in your space whose posts you want to reply to, and the questions your buyers ask on X. ### Week 2: find your rhythm 1. Post three times: one progress post (what shipped), one lesson (what you learned, with a number if you have one), one decision (what you’re choosing between, and asking for input). 2. Reply to five posts a day where you can add something. Answer first; mention your product only where it fits. 3. Keep each post to one idea. A post holds up to 280 characters, and a link counts as 23 of them, whatever its length. ### Week 3: show the work 1. Share one real metric and what you’ll do about it. Specific beats impressive. 2. Post a short screen recording of a feature in use. 3. Tag every link you share with UTM parameters (utm_source, utm_medium, utm_campaign) so you can see which posts bring visitors. 4. If teammates post about the product, ask them to say they work on it. The FTC’s guidance says employees should disclose their relationship, and a profile line isn’t enough. ### Week 4: launch 1. Tell the launch as a story: the problem, what you tried, what you built, and what you want people to try. 2. If your product is something people can try, post a Show HN on Hacker News. Its rules say Show HN is for something you’ve made that people can play with, not landing pages or sign-up pages, and ask you not to recruit upvotes or comments from friends. 3. Reply to every comment on launch day. 4. Look back: which posts brought visitors and sign-ups? Do more of those next month. ### How Odeist AI helps Odeist AI’s X posting agent writes posts in your voice on the schedule you set, and its X reply agent finds posts about your space and drafts replies you approve, edit or skip. For launch week, the launch kit writes the X thread, a LinkedIn post, an email, a press release and a plan, and the Hacker News agent drafts your Show HN for you to post yourself. ### What if I have no numbers to share yet? Share decisions and lessons instead. “We cut onboarding from 6 steps to 3, here’s why” is building in public without a revenue chart. ### Should every post link to my product? No. Most posts should stand on their own. Put the link in your profile and use it in a post when the post is about something people can try. ### Sources 1. X developer platform: Counting characters: https://docs.x.com/fundamentals/counting-characters (read 2026-11-06) 2. Google Analytics Help: URL builders: Collect campaign data with custom URLs: https://support.google.com/analytics/answer/10917952 (read 2026-11-06) 3. FTC: FTC’s Endorsement Guides: What People Are Asking: https://www.ftc.gov/business-guidance/resources/ftcs-endorsement-guides-what-people-are-asking (read 2026-11-06) 4. Hacker News: Show HN guidelines: https://news.ycombinator.com/showhn.html (read 2026-11-06) ## LinkedIn for founders: a weekly routine Source: https://odeist.com/learn/linkedin-for-founders-playbook (updated 2026-11-06) Post twice a week about what you know from building your company, spend fifteen minutes a day commenting thoughtfully on posts from people in your market, and skip the tricks. LinkedIn says its feed favours professionally relevant posts with genuine insight and is acting against engagement bait, pods and comment automation. ### What LinkedIn says it rewards LinkedIn’s help pages say that for recommended content it considers signals such as whether the content is high quality and professionally relevant. In March 2026 it said members will see more posts that offer genuine insight, actionable ideas and thoughtful perspectives, and fewer designed purely to game distribution. It also said it is reducing engagement bait such as “comment to agree” prompts, working to make engagement pods ineffective, and curbing comment automation. Its Professional Community Policies ask members not to artificially increase engagement or agree ahead of time to like or re-share each other’s posts. ### The weekly routine 1. Monday: write down one thing you learned last week from customers, sales or building. That is Tuesday’s post. 2. Tuesday: post it. One idea, your own experience, a clear takeaway. No “comment YES” endings. 3. Every day: fifteen minutes commenting on posts from buyers and peers. Add a real point or a question; skip “Great post!”. 4. Thursday: post a practical one: a checklist, a before-and-after, or a mistake and its fix. 5. Friday: answer every comment on your posts and note which post started real conversations. ### Turn conversations into customers - When someone asks a question you can answer, answer it in the comments first. - Don’t use connection invitations to pitch people you don’t know: LinkedIn’s policies don’t allow it. - If teammates share your posts or praise the product, they should mention they work for you. ### How Odeist AI helps Odeist AI’s LinkedIn posting agent writes posts in your voice from your brand profile on the schedule you choose, and its LinkedIn reply agent finds posts about your space and drafts comments. With approvals on, each draft waits for you to approve, edit or skip, which keeps a person behind every comment. ### Should I post from my personal profile or my company page? We suggest your personal profile for conversations and your company page for announcements, which you then share from your profile. ### Is it fine to use AI to write LinkedIn posts? LinkedIn’s policies ask for original, professional, relevant content and authentic engagement. Use AI to draft from your own experience, and read and edit every post and comment before it goes out. ### Sources 1. LinkedIn Help: LinkedIn Feed, overview: https://www.linkedin.com/help/linkedin/answer/a523360 (read 2026-11-06) 2. LinkedIn: How LinkedIn Is Improving the Feed to Show More Relevant, Authentic Professional Content (12 March 2026): https://news.linkedin.com/2026/ImprovingTheFeed (read 2026-11-06) 3. LinkedIn Professional Community Policies: https://www.linkedin.com/legal/professional-community-policies (read 2026-11-06) 4. FTC: FTC’s Endorsement Guides: What People Are Asking: https://www.ftc.gov/business-guidance/resources/ftcs-endorsement-guides-what-people-are-asking (read 2026-11-06) ## Marketing glossary Source: https://odeist.com/learn/glossary - AI crawler: An AI crawler is a bot that fetches web pages for an AI company, either to train models or to answer people’s questions with search. The two jobs usually have different bots, and you can allow one and block the other. (https://odeist.com/learn/glossary/ai-crawler) - AI Overviews and AI Mode: AI Overviews and AI Mode are Google Search features that answer a query with an AI-written response and links to pages that support it. (https://odeist.com/learn/glossary/ai-overviews) - AI share of voice: AI share of voice is how often AI assistants name your brand, compared with your competitors, when they answer the questions your buyers ask. (https://odeist.com/learn/glossary/ai-share-of-voice) - Answer engine optimization (AEO): Answer engine optimization (AEO) is writing and structuring pages so a search engine or AI assistant can lift a direct answer from them. In practice it overlaps almost completely with GEO. (https://odeist.com/learn/glossary/answer-engine-optimization) - Brand voice: Brand voice is the consistent personality in everything your business writes: the words you choose, how formal you are, whether you joke, and what you never say. (https://odeist.com/learn/glossary/brand-voice) - Building in public: Building in public is sharing how you build your company as you go: what you shipped, what you learned, your numbers and your mistakes, usually on X or LinkedIn. (https://odeist.com/learn/glossary/building-in-public) - Buying signal: A buying signal is a public post or action that shows someone is looking for what you sell right now: asking for a recommendation, comparing two tools, or complaining about the one they use. (https://odeist.com/learn/glossary/buying-signal) - Citation in an AI answer: A citation in an AI answer is a link the assistant shows as a source for what it said. Being cited sends readers to your page; being named without a link still shapes what buyers think. (https://odeist.com/learn/glossary/ai-citation) - Credits (in AI marketing tools): Credits are a single unit that an AI tool charges for its work: each job, such as a post, a reply or an article, costs a set number of credits from one balance. (https://odeist.com/learn/glossary/marketing-credits) - E-E-A-T: E-E-A-T stands for experience, expertise, authoritativeness and trustworthiness: the qualities Google’s systems look for when deciding which helpful, reliable content to show. (https://odeist.com/learn/glossary/e-e-a-t) - Engagement bait: Engagement bait is a post written to trigger likes, comments or shares for their own sake, such as “comment YES if you agree”, rather than to say something useful. (https://odeist.com/learn/glossary/engagement-bait) - Engagement pod: An engagement pod is a group of people who agree to like and comment on each other’s posts to push them up the feed. (https://odeist.com/learn/glossary/engagement-pod) - Generative engine optimization (GEO): Generative engine optimization (GEO) is the work of getting your business named, and your pages cited, in answers written by AI assistants such as ChatGPT, Gemini, Claude and Perplexity. (https://odeist.com/learn/glossary/generative-engine-optimization) - Google Search Console: Google Search Console is Google’s free service for seeing how your site appears in Google Search: which searches show it, how often people click, and what stops pages from being indexed. (https://odeist.com/learn/glossary/google-search-console) - Human in the loop (approvals): Human in the loop means a person reviews what an AI tool wrote before it is published. In marketing, it usually means approving, editing or skipping each draft post and reply. (https://odeist.com/learn/glossary/human-in-the-loop) - llms.txt: llms.txt is a proposed Markdown file at the root of a website that gives AI tools a short summary of the site and links to its most useful pages. (https://odeist.com/learn/glossary/llms-txt) - robots.txt: robots.txt is a plain text file at the top of a website that tells crawlers which paths they may and may not fetch. (https://odeist.com/learn/glossary/robots-txt) - Scaled content abuse: Scaled content abuse is Google’s name for publishing many pages mainly to manipulate search rankings rather than to help readers, no matter how the pages are made. (https://odeist.com/learn/glossary/scaled-content-abuse) - Share of voice: Share of voice is your part of the conversation in your market: the mentions, posts or answers that name you, divided by those that name you or a competitor. (https://odeist.com/learn/glossary/share-of-voice) - Show HN: Show HN is a section of Hacker News for sharing something you made that other people can try. (https://odeist.com/learn/glossary/show-hn) - Social listening: Social listening is watching public conversations for your brand, your competitors and the problems you solve, then acting on what you find: answering a question, fixing a complaint, or learning how buyers talk. (https://odeist.com/learn/glossary/social-listening) - Structured data (schema markup): Structured data is code on a page, usually JSON-LD using the schema.org vocabulary, that tells search engines what the page is about: an organization, a product, an article, a set of questions and answers. (https://odeist.com/learn/glossary/structured-data) - Tone of voice: Tone of voice is how your brand voice adapts to a moment: lighter in a launch post, plainer in a reply to a frustrated customer. (https://odeist.com/learn/glossary/tone-of-voice) - UTM parameters: UTM parameters are tags added to the end of a link, such as ?utm_source=linkedin, that tell your analytics where a visit came from. (https://odeist.com/learn/glossary/utm-parameters) ## Which AI marketing tools do ChatGPT and Claude recommend? Source: https://odeist.com/blog/ai-marketing-tools-in-ai-answers-2026-11 (updated 2026-11-19) Short answer: Products AI answers name most: ChatGPT came out on top at 38%, in Odeist AI's data from 48+ answers (19 November 2026). In short - Products AI answers name most: ChatGPT had the highest share of answers naming the product (38%) and Profound the lowest (13%), across 12 groups of at least 48 answers each. - AI marketing for small teams: the products named most: HubSpot had the highest share of answers naming the product (69%) and Jasper the lowest (38%), across 5 groups of at least 16 answers each. - Posting and scheduling with AI: the products named most: Hootsuite had the highest share of answers naming the product (56%) and ChatGPT the lowest (25%), across 5 groups of at least 16 answers each. - Listening and AI-answer tracking: the products named most: Peec AI had the highest share of answers naming the product (50%) and Ahrefs the lowest (31%), across 5 groups of at least 16 answers each. - Products named per answer, by engine: Claude had the highest products named per answer (8.9) and ChatGPT the lowest (5), across 2 groups of at least 24 answers each. - Sites the answers cite most: buffer.com had the highest share of answers citing the site (15%) and ploy.ai the lowest (4.2%), across 8 groups of at least 48 answers each. We asked ChatGPT and Claude 24 questions buyers ask about AI marketing tools, with web search on, and counted the products each of the 48 answers named. Odeist AI made this study and is one of the products in this market: it is counted like every other product, from the same answers. ### What we found #### Products AI answers name most GroupShare of answers naming the productAnswersChatGPT38%48+Buffer35%48+Hootsuite25%48+HubSpot23%48+Sprout Social21%48+Canva19%48+Jasper17%48+Peec AI17%48+Semrush17%48+Brand2415%48+Claude15%48+Profound13%48+ Of the 48 answers, the share that named each product at least once. #### AI marketing for small teams: the products named most GroupShare of answers naming the productAnswersHubSpot69%16+ChatGPT63%16+Buffer56%16+Canva50%16+Jasper38%16+ 16 answers to 8 questions. #### Posting and scheduling with AI: the products named most GroupShare of answers naming the productAnswersHootsuite56%16+Buffer50%16+Sprout Social44%16+Vista Social31%16+ChatGPT25%16+ 16 answers to 8 questions. #### Listening and AI-answer tracking: the products named most GroupShare of answers naming the productAnswersPeec AI50%16+Brand2444%16+Semrush44%16+Profound38%16+Ahrefs31%16+ 16 answers to 8 questions. #### Products named per answer, by engine GroupProducts named per answerAnswersClaude8.924+ChatGPT524+ How many different products an answer names, on average. #### Sites the answers cite most GroupShare of answers citing the siteAnswersbuffer.com15%48+semrush.com10%48+hootsuite.com8.3%48+sproutsocial.com8.3%48+apaya.com6.3%48+brand24.com6.3%48+canva.com6.3%48+ploy.ai4.2%48+ Of the answers, the share that cited each site at least once. ### How we made it - 24 questions in three groups (AI marketing for small teams, posting and scheduling with AI, listening and AI-answer tracking), written the way buyers ask them; none names a product. They are listed below. - Each question went to each engine once, with its web search on, between 2026-11-19 and 2026-11-19: Claude (claude-haiku-5-5), ChatGPT (gpt-5.4-mini-2026-03-17). - A product counts once per answer, however often the answer names it. Names are read from the answer text; a small model then lists the products each answer recommends, and a name it returns is kept only if it is in the text. - 48 of 96 requests failed (grok, gemini: the provider account was out of credits) and are left out. - AI answers change from day to day and differ by account and place: this is one reading, dated. - The questions: "What are the best AI marketing tools for a startup with no marketing hire?"; "Which AI tool can run social media marketing for a small SaaS company?"; "Best AI marketing assistant for solo founders in 2026?"; "What AI agents can find customers for a B2B startup?"; "Cheapest AI tool to replace a part-time marketing freelancer?"; "Which AI marketing platforms are good for technical founders who hate marketing?"; "Best AI tools to get the first 100 users for a new app?"; "What is the best all-in-one AI marketing tool for a two-person company?"; "What are the best AI tools to write and schedule LinkedIn and X posts?"; "Best social media scheduler with AI writing in 2026?"; "Which tool posts to X, LinkedIn, Instagram and Threads with AI-written drafts?"; "Best open-source social media scheduling tool?"; "What tool writes social posts in my brand voice and lets me approve them first?"; "Best AI tool to repurpose blog posts into social media posts?"; "Which social media management tools are best for agencies with many clients?"; "Best Buffer or Hootsuite alternative with AI?"; "What are the best social listening tools for startups?"; "Best tool to track mentions of my brand on Reddit, X and Hacker News?"; "Which tools find people asking for a product like mine on Reddit and X?"; "Best tools to track whether ChatGPT and Gemini recommend my product?"; "How can I monitor my brand in AI search answers, and which tools do it?"; "Best affordable alternative to Brand24 or Mention for brand monitoring?"; "Which tools show my share of voice against competitors in AI answers?"; "Best tools for generative engine optimization (GEO) in 2026?". - Published 19 November 2026. Groups with fewer than 1 answers are not shown. ### Sources - [The answers, products named and sources cited, for every question and engine (JSON)](https://odeist.com/ai-marketing-tools-in-ai-answers-2026-11.json) (19 November 2026) - [OpenAI Responses API with web search](https://platform.openai.com/docs/guides/tools-web-search) (19 November 2026) - [Gemini API grounding with Google Search](https://ai.google.dev/gemini-api/docs/google-search) (19 November 2026) - [Anthropic web search tool](https://docs.anthropic.com/en/docs/agents-and-tools/tool-use/web-search-tool) (19 November 2026) - [xAI live search](https://docs.x.ai/docs/guides/live-search) (19 November 2026) ### Use this chart The chart is free to use on your site with a link back to this page. Paste this:
Products AI answers name most. ChatGPT 38%, Buffer 35%, Hootsuite 25%, HubSpot 23%, Sprout Social 21%, Canva 19%, Jasper 17%, Peec AI 17%, Semrush 17%, Brand24 15%, Claude 15%, Profound 13%.
Source: Odeist AI, AI marketing tools in AI answers (19 November 2026)
### Where does this data come from? The answers, products named and sources cited, for every question and engine (JSON). It covers 19 November 2026. ### Can anyone in the data be identified? No. Every number is a median over a group of at least 1 answers, rounded, with no names or post text. Smaller groups were left out. ### Can I use the chart? Yes. Paste the embed code from this page, or use the image with a link to https://odeist.com/blog/ai-marketing-tools-in-ai-answers-2026-11. ## Turn a Photo or a Voice Note Into a LinkedIn Post in Two Minutes Source: https://odeist.com/blog/photo-voice-note-to-linkedin-post (updated 2026-11-16) Short answer: capture the idea the moment you have it: a photo, or thirty seconds of voice saying what happened and why it matters. Turn it into a draft in your brand's voice, read it once, ask for one change at a time, and post. You never sit down to write from a blank page. In short - The best LinkedIn posts start as moments, not as writing sessions. - A photo needs one sentence of context: what happened and why it matters. - A voice note should answer three things: what, so what, now what. - Read the draft for names and numbers, then ask for one change at a time. ### Why moments make better posts than writing sessions Sit down to "write a LinkedIn post" and you get a generic post. The posts people stop scrolling for are specific: a customer's exact words, a whiteboard from a planning session, the day the team shipped something hard. Those moments happen away from your desk, and they fade fast. So the job is capture, not writing. ### Turning a photo into a post A photo on its own is a postcard. A photo with a reason is a story. When you send one, add a line: - "Our team at the Pune design meetup. Three customers came to say hi." - "The whiteboard from today's pricing session. We're dropping the setup fee." - "Our first order shipped to Canada." That one line is enough for a good draft: the moment, why it matters, and a question or a takeaway to end on. Photos of real people and real places do better than polished graphics, and you already have them. ### Turning a voice note into a post Talking is faster than typing, and it sounds more like you. Before you record, think of three questions: - What happened? The fact, in one or two sentences. - So what? Why it matters to the people who follow you. - Now what? What you'd like them to think, try or ask. Thirty to sixty seconds is plenty. Don't worry about ums and restarts; the words are written out first, then shaped into a post. Keep notes under two minutes; long ones make for long, unfocused drafts. ### Reading the draft The first draft is a draft. Read it once for: - Names and numbers. Are they right? Did it invent anything? - Claims. Would you say this out loud to a customer? - The first line. Would you stop scrolling for it? Then ask for one change at a time: "shorter", "start with what the customer said", "less formal", "end with a question". Each comes back as a new version, so you can stop the moment it's right. ### A two-minute routine - At the event, take the photo. - On the way out, send it with a sentence, or record a voice note. - A minute later, the draft is in your chat. - Read it, ask for one change, approve. ### Doing this with Odeist AI In Odeist AI, you do all of this inside WhatsApp, Telegram or Slack. Send the photo or the voice note to Odeist AI, say where it should go ("LinkedIn", "X and LinkedIn"), and your brand's posting agent writes the draft in your voice, with your photo. It comes back as a card with Approve, Edit and Skip. Nothing is posted until you tap Approve or say yes. Each post uses your plan's credits, about 30 for a post with an image. A voice note counts as one message to Odeist AI. ### Can I write a LinkedIn post from a voice note? Yes. Record what you want to say, have it written out, and turn it into a post in your brand’s voice. In Odeist AI you send the voice note in WhatsApp, Telegram or Slack and the draft comes back in the same chat. ### What makes a photo work on LinkedIn? Real people and real moments: your team, a customer, a whiteboard, an event. A short caption that says what happened and why it matters does more than a polished stock image. ### Should I post the first draft as it is? Read it first. Check names, numbers and anything you would not say yourself. Then ask for one change at a time, such as a shorter opening or a clearer last line. ## How to Get Your Brand Recommended by ChatGPT and Other AI Assistants Source: https://odeist.com/blog/get-recommended-by-chatgpt (updated 2026-11-05) Short answer: AI assistants recommend brands they can read, understand and see confirmed elsewhere. Say plainly on your own site what you do and for whom, answer the questions buyers ask in full sentences, get mentioned on sites and communities your buyers trust, and check a fixed list of questions in the assistants every month. In short - Write for a reader who has never heard of you: what it is, who it's for, what it costs. - Answer real buyer questions on your site, in plain text an assistant can quote. - Be mentioned next to the problem you solve on other sites, not only on yours. - Measure with the same questions every month, across several assistants. ### How AI assistants choose what to recommend Nobody outside the AI companies knows the exact rules, and they change. But the behaviour is consistent enough to work with. When someone asks "what's a good invoicing tool for freelance designers?", an assistant answers from what it has read on the web, and sometimes from a live search. It tends to name products that: - say clearly what they are and who they're for, - match the question's words: the problem, the audience, the use case, - appear in more than one place, so the claim is confirmed by someone else. None of this is new. It's good marketing, written down where machines can read it. ### 1. Make your site easy to summarise An assistant has to turn your site into one sentence. Help it: - First paragraph of your homepage: what the product is, for whom, and the main result, in plain words. "Acme is invoicing software for freelance designers that chases late payments for you" beats "Get paid, beautifully." - A page per audience or use case if you serve several. One page that tries to speak to everyone gets summarised as no one. - Pricing in text, not only in an image or behind a sales call. Price is one of the first things people ask an assistant. - Plain HTML text for the important claims. Text inside images, sliders or scripts that load late is easy to miss. ### 2. Answer the questions buyers actually ask Collect the questions from sales calls, support tickets and community threads, then answer them on your site: Question typeExampleWhere to answer itFit"Is it good for a team of three?"An FAQ or a page for that audienceComparison"How is it different from X?"An honest comparison pageHow-to"How do I chase a late invoice politely?"A guide on your blogCost"How much does it cost a month?"The pricing page, in text Lead each answer with the answer, then the detail. That is the part an assistant is most likely to quote. ### 3. Be mentioned where your buyers already look Your own site says you're good. Other sites saying it is what makes a recommendation feel safe. Useful places: - Review and directory sites in your category. - Communities such as Reddit, Hacker News or industry forums, through genuinely useful answers, never spam. - Articles and lists in your niche: guest posts, podcasts, "tools we use" pages from customers. - Consistent naming everywhere: the same product name and the same one-line description. ### 4. Add the small technical signals - Structured data (schema.org) for your organisation, product and FAQs. - An llms.txt file, a plain list of your most useful pages for AI tools. It's a proposal, not a standard every assistant uses, but it takes ten minutes. Odeist AI has a free generator for it. - A sitemap and fast pages, which help classic search as much as AI search. ### 5. Measure it the same way every month Pick ten to twenty questions a buyer would ask, word them the way a buyer would, and ask them in several assistants once a month. Note whether you're named, in what position, and what's said about you. Watch the trend, not a single answer: assistants vary from one try to the next. In Odeist AI, the AI answers agent asks your questions for you on a schedule, shows where you and your competitors appear, and leaves the fixes it finds in Today for you to approve. ### How does ChatGPT decide which brands to recommend? No one outside the AI companies knows the exact rules. What is clear is that assistants lean on what they can read on the web: pages that say plainly what a product does and for whom, and other sites that mention it in the same breath as the problem. ### Is AI search different from SEO? It overlaps a lot. Clear pages, good answers to real questions and mentions on trusted sites help both. The difference is that an assistant gives one short answer instead of ten links, so being specific about who you serve matters even more. ### How long does it take to show up in AI answers? It varies with how often each assistant refreshes what it reads. Treat it as a monthly check, not a daily one, and track a fixed set of questions so you can see the trend. ### What is llms.txt? A plain text file at the root of your site that lists your most useful pages for AI tools, in the way a sitemap does for search engines. It is a proposal, not a standard every assistant uses, but it is cheap to add. ## How to Approve Social Media Posts From WhatsApp (Without Opening Another App) Source: https://odeist.com/blog/approve-social-posts-from-whatsapp (updated 2026-11-02) Short answer: to approve social posts from WhatsApp, have each draft arrive in the chat as a card with its picture, its words and three buttons: approve, edit, skip. Reply in plain words when you want a change ("shorter", "post it tomorrow at 9"). Let writing and rewriting happen straight away, and make anything that publishes wait for your yes. In short - Drafts that wait for a dashboard visit wait for days. Drafts that arrive where you already are get answered. - A card should show exactly what will go out: picture, words, network. - Edits work best in your own words, one change at a time. - Rewriting can happen at once. Posting, scheduling and skipping should ask first. ### Why approvals are where posting stalls Most small teams don't run out of ideas. They run out of moments. A draft is ready on Tuesday morning, you're in meetings until Thursday, and by then the post about Tuesday's launch is old news. The fix isn't more reminders to open another app. It's putting the decision where you already spend your day: your phone's messages. ### What a good approval message looks like A notification that says "You have 3 posts to review" and a link is still a dashboard visit, only with an extra tap. A good approval message is the draft itself: - The picture or clip the post will carry, at the top. - The words, exactly as they'll be posted. - Where it goes: LinkedIn, X or Instagram, and for which brand if you run more than one. - When it goes, if it's scheduled. - Three buttons: Approve, Edit, Skip. If you can decide from the message alone, you'll decide in seconds. ### Editing in plain words The edit step is where most tools send you back to a form. In a chat you can just say what you want: You writeWhat should happen"shorter"The same post, tighter, back in the chat as a new card."start with the customer's problem"A new opening line, the rest kept."add our website at the end"The link added, nothing else changed."post it tomorrow at 9"A question: schedule for tomorrow 9:00 your time? Yes or No."skip this one, it's off topic"A question first, then the draft is dropped and the reason teaches the agent. One change at a time works better than a list. You see each version and stop when it's right. ### What should always ask first A simple rule keeps chat approvals safe: anything you can undo can happen at once; anything public waits for a yes. - At once: writing a draft, rewriting a draft, answering a question. Nothing leaves your account. - Asks first: approving (it posts), scheduling (it will post), skipping, running an agent that spends credits, closing a task. A button tap on the draft's own card is your yes. A typed "yes" works too. Anything longer than a yes or a no is read as a new message, so "yes but shorter" gets you a shorter draft, not a post. ### Starting a post from the chat Approvals are half of it. The other half is starting posts when the idea comes: - Send a photo from an event and say "post this on LinkedIn". - Send a voice note with what you want to say; it's written out and turned into a draft. - Or just type "write an X post about our new pricing page". The draft comes back in the same chat a minute later, as a card, ready for the same three buttons. ### The 24-hour rule on WhatsApp WhatsApp lets a business message you freely for 24 hours after your last message. After that, it allows only short pre-approved notices. So if you haven't written in a day, Odeist AI sends a brief "drafts are ready" notice; one tap on it opens the conversation again and the cards follow. ### Setting it up in Odeist AI - Add your WhatsApp number in Settings › Profile. - Under Settings › Chat & notifications › Approve drafts in, pick WhatsApp. Telegram and Slack work the same way once connected; the review page by link is still there if you prefer it. - Say hello to Odeist AI on WhatsApp. The next drafts your agents write will arrive as cards. You can try it on the 7-day free trial; plans start at $29 a month. ### Can I approve social media posts from WhatsApp? Yes, if your tool sends drafts into a WhatsApp chat. In Odeist AI each draft arrives as a card with its image and words, and buttons to approve, edit or skip it. You can also reply in plain words, like “shorter” or “post it tomorrow at 9”. ### Is it safe to let a chat app post for me? It is when nothing goes out without a clear yes from you. Writing and rewriting a draft can happen straight away, because nothing is published. Posting, scheduling and skipping should always ask first. ### What happens if I don’t reply for a day? WhatsApp only lets a business start a conversation with an approved template after 24 hours of silence. Odeist AI then sends a short notice; one tap opens the chat again and shows the drafts waiting. ### Does it work on Telegram and Slack too? Yes. You choose where drafts arrive under Settings › Chat & notifications › Approve drafts in: WhatsApp, Telegram, Slack, or the review page by link. ## Best AI Answer Tracking Tools for Startups in 2026, Prices Checked Source: https://odeist.com/blog/ai-answer-tracking-tools (updated 2026-11-06) Short answer: most startups should check by hand first: ask 15 buyer questions in ChatGPT, Gemini, Claude and Perplexity once a month and log who gets named. When that becomes a chore, Otterly AI (from $29 a month) and LLMrefs ($79 a month) are affordable dedicated trackers; Peec AI suits teams that want daily data, and Profound is built for enterprises (prices as of November 2026). If you want the tracking and the fixes in one place, Odeist AI's AI answers agent checks ChatGPT, Gemini, Claude and Grok and hands you what to change. In short - Track mostly the questions buyers ask before they know your name, not searches for your brand. - Measure five things: named or not, position, what is said, who is named instead, and which pages get cited. - A spreadsheet and two hours a month is a real tracking system. Start there. - Pay for a tool when you need more prompts, more assistants or a weekly rhythm. - A tracker only shows the gap. Budget time, or a tool, for the fixes. ### Why does tracking AI answers matter for a startup? When a buyer asks ChatGPT "what's a good bookkeeping tool for freelancers?", they get a short list, not ten blue links. If you're not on it, you're not in the running, and nothing in your analytics will tell you. Tracking tells you three things you can act on: whether you're on the list, who is on it instead, and which pages the assistant read to decide. The third one is where most of the fixing happens. ### What should you measure? Ignore vanity scores until you understand the raw answers. For each question and each assistant, write down: What to logWhy it mattersNamed or notThe basic signal. Count it as a share of your questions.Position in the listFirst of five and fifth of five are different results.What it says about youWrong prices, old features or the wrong audience cost you buyers who did hear your name.Who is named insteadYour real competitors in AI answers, which are not always the ones you expect.Which pages it citesThe review sites, lists and threads you need to be on. ### Which questions should you track? Group your questions into four kinds, because each kind fails for a different reason. Say you sell bookkeeping software for freelancers: KindSample questionIf you're missing, it usually meansBest-of"best bookkeeping app for freelancers"You're not on the lists and review sites the assistant reads.Job to be done"how do I track expenses as a freelance designer"Your site explains the product, not the problem.Competitor"alternatives to [big competitor] for freelancers"No comparison page, and nobody else compares you.Branded"is [your product] good for freelancers"Your own pages are thin, out of date or hard to read. Fifteen questions is plenty to start: five best-of, five job to be done, three competitor, two branded. Word them the way a buyer would type them, typos and all, not the way your homepage talks. ### How do you check for free, by hand? You can do this today with no tool at all. - Write your 15 questions in a spreadsheet, one per row. - Add a column per assistant: ChatGPT, Gemini, Claude and Perplexity cover most of what buyers use. - Ask each question in a fresh chat, logged out or with memory off where you can, so your own history doesn't nudge the answer. - Ask each one twice. In our experience the same question can get a different list on a second try. If you're named once out of two, mark it as a half. - Log the five things from the table above. Paste the cited links into a notes column. - Repeat on the same day each month. Watch the trend across all your questions, not any single answer. That's 15 questions, four assistants, twice each: about two hours a month. If you only want a one-off snapshot first, HubSpot's free AI Search Grader gives a one-time score for how ChatGPT, Perplexity and Gemini see your brand, without creating an account. Two hours is fine for a month or two. It stops being fine when you want weekly checks, more questions, or answers for another country or language. ### What are the best AI answer tracking tools in 2026? We checked each tool's own pricing page. Prices are as of November 2026, in US dollars, on monthly billing unless noted. They change often, so check before you buy. ToolStarting priceAssistants coveredBest fitHubSpot AI Search GraderFreeChatGPT, Perplexity, GeminiA one-off snapshot before you commit to anything.Otterly AI$29 a month (Lite, 15 prompts)ChatGPT, Google AI Overviews, Perplexity, Microsoft Copilot; Gemini, AI Mode and Claude as paid add-onsThe smallest budget that still tracks every week.LLMrefs$79 a month (500 prompts, marked as a limited-time price)ChatGPT, Gemini, Perplexity, Claude, Grok, Copilot, Meta, AI Overviews, AI ModeThe most assistants and prompts for the money.Peec AI$95 a month (Starter, 50 prompts, 3 active models, daily)ChatGPT, Gemini, Copilot, AI Overviews, AI Mode, Naver AI; Perplexity and Claude on EnterpriseMarketing teams that want daily data and clean reporting.Ahrefs Brand RadarFrom $199 a month as an add-on; paid Ahrefs plans (Lite at $129) include some AI prompt checksChatGPT, Gemini, Perplexity, Copilot, AI Overviews, AI Mode; Claude for custom promptsTeams already using Ahrefs for SEO.ProfoundFree 7-day trial (50 preset prompts), then Enterprise pricing on requestTrial: ChatGPT, Gemini, AI Overviews. Enterprise: up to 10, including Claude, Perplexity and CopilotLarger companies with a budget and a team to read the data.Odeist AIPlans from $29 a month; each answer uses creditsChatGPT, Gemini, Claude, GrokSmall teams that want the fixes done, not only measured. A few honest notes on the table: - Otterly AI is the easiest paid start. The add-on prices for Gemini and Claude matter: if your buyers use Gemini, budget for it. - LLMrefs covers the widest set of assistants in one price. Check how prompts and keywords are counted on its page before you plan around 500. - Peec AI lets Starter run 3 active models, so pick the three your buyers use. - Profound lists only its free trial and Enterprise on its pricing page, so treat it as an enterprise tool unless you're ready for a sales call. - Odeist AI doesn't check Perplexity or Google's AI Overviews today. If those matter most to your buyers, one of the tools above fits better for that part. ### How do you choose the right one? Answer three questions: - Which assistants do your buyers use? A developer tool and a consumer app get very different answers. Ask five customers which one they asked last, and pick a tool that covers those. - How many questions do you need? Fifteen to fifty is enough for most startups to start with. Paying for hundreds before you've fixed the first fifteen is money spent on charts. - Who does the fixing? Every tracker shows the gap. Few close it. If nobody on your team has a free afternoon a week for pages, comparisons and outreach, the cheapest tracker becomes a dashboard nobody opens. ### What do you do once you see the gaps? Tracking is the easy half. Here is what each gap usually calls for: - Missing on best-of questions: look at the pages the assistant cited. Those lists, directories and review sites are your outreach list. - Missing on job-to-be-done questions: write a page or article that answers that exact question, answer first, in your buyer's words. - Wrong facts about you: fix your pricing and features in plain text on your own site, then check whether the wrong claim comes from an old third-party page you can ask to update. - A competitor always named first: read the pages cited next to them. You'll usually find a comparison or a list you're not on. Our guide on [how to get recommended by ChatGPT and other AI assistants](https://odeist.com/blog/get-recommended-by-chatgpt) goes through each of these fixes step by step. This is where Odeist AI is different from a tracker. Its AI answers agent asks your buyers' questions on a schedule, sees who the assistants recommend, and leaves fixes for your site in Today, plus article ideas with a Write an article button. A gap in AI answers can also become a creator pitch to writers the assistants already cite, and that pitch always waits for your approval. ### Track it and fix it with Odeist AI Odeist AI is a team of AI agents for startups and small teams. For AI answers, it gives you: - The AI answers agent: checks what ChatGPT, Gemini, Claude and Grok say when buyers ask about your category, for the country and language you set, with its findings in Analytics. - The SEO agent: a daily audit of your site with ready fixes, applied to your CMS or opened as a GitHub pull request. - The Articles agent: writes the articles your gaps call for and publishes them to WordPress, Webflow, Sanity or GitHub. Each article waits for you unless you turn on automatic publishing. Odeist AI prices by credits, not by prompt bundles. Every answer uses credits, and the cost differs by assistant (each one is on the [pricing page](https://odeist.com/pricing)), so your monthly cost depends on how many questions you track, on which assistants, and how often. You choose the assistants in the agent's settings, and Settings › Credits shows what the agent uses in a month at those settings. If credits are new to you, how usage-based pricing works explains them in plain words. Every plan has every agent, so the same credits also run your SEO audits, articles and replies. Each agent has one switch: Ask me first sends you every draft to approve, in the app, WhatsApp, Telegram or Slack. On its own lets it run. If you want a second opinion on where that line should sit, read what an AI marketing agent should do without you. Growth agents like AI answers come with a plan or the trial. The trial is 7 days of Growth with 1,000 credits, and you don't need a card. [Start free](https://odeist.com/signup), or first [get a free growth report](https://odeist.com/audit) on your website: see what your marketing is missing, free, in about a minute. Whichever tool you pick, the habit is what moves the number: the same questions, every month, and one fix shipped for every gap you find. ### What is an AI answer tracking tool? A tool that asks AI assistants such as ChatGPT, Gemini, Claude or Perplexity the questions your buyers ask, on a schedule, and records whether your brand is named, in what position, what is said and which pages are cited. ### How do I check if ChatGPT mentions my startup for free? Write 15 questions a buyer would ask, ask each one twice in a fresh ChatGPT chat with memory off, and log whether you are named, your position, who is named instead and the cited pages. Repeat monthly in Gemini, Claude and Perplexity too. ### What is the cheapest AI answer tracking tool? HubSpot's AI Search Grader is free for a one-time check. Among the paid trackers we checked, Otterly AI starts at $29 a month for 15 prompts, as of November 2026, with Gemini and Claude as paid add-ons. ### Which assistants does Odeist AI track? ChatGPT, Gemini, Claude and Grok; you choose which in the AI answers agent's settings. It does not check Perplexity or Google's AI Overviews today. ### How much does AI answer tracking cost in Odeist AI? Plans start at $29 a month, and every answer uses credits at a cost that differs by assistant, listed on the pricing page. You can try it free for 7 days on Growth with 1,000 credits, no card needed. ## Cheap Alternative to a Social Media Manager: 5 Options, Real Prices Source: https://odeist.com/blog/cheap-alternative-social-media-manager (updated 2026-11-06) Short answer: the cheapest real alternative to a social media manager for a two-person startup is to split the job. Let a tool do the repeat work and have one founder spend a short, fixed slot each day approving and replying. A scheduler only publishes what you write. A freelancer or agency covers more, but costs hundreds to thousands of dollars a month. An AI team like Odeist AI posts and finds people to reply to, from $29 a month. In short - A social media manager does five jobs. Most cheap options cover one or two of them. - Schedulers are the cheapest tool, but they don't find people or write replies. - Freelancers bill by the hour, and the replies that win customers need hours of product knowledge. - Agencies are built for brands with budgets in the thousands a month. - Choose by the job you're missing, not by the lowest price. ### What does a social media manager actually do? Before you replace someone, list what you'd be replacing. A good social media manager does five jobs: - Plans and writes posts in a voice that sounds like the company. - Publishes them on time, on each network. - Finds conversations where people describe the problem you solve, including people who don't follow you yet. - Replies, both to those conversations and to anyone who comments. - Reports on what worked, and changes the plan. Most cheap options cover jobs 1 and 2, because those are the visible ones. At two people, jobs 3 and 4 usually matter more. A new account has few followers, so your posts reach few people. A useful reply in someone else's thread reaches everyone reading it. We made that case in [how to get customers from X replies](https://odeist.com/blog/get-customers-from-x-replies). So the question isn't "what's the cheapest way to post?". It's "which of the five jobs can we not afford to drop?". ### What does each alternative cost, and what does it cover? Third-party prices are as of November 2026, from each company's own pages. OptionWhat it costsWrites postsFinds people to reply toWrites repliesDo it yourself$0, plus your hoursYou doYou do, by handYou doScheduler (Buffer, Hootsuite)Buffer: free for 3 channels, then from $5 a channel a month billed yearly. Hootsuite: from $99 a monthNo, it publishes what you writeNoNoFreelancer (Upwork)Upwork median $20 an hour, most between $14 and $35YesOnly if you pay for those hoursYes, within their hoursAgency (WebFX)From $3,150 a monthYes, plus designDepends on the contractDepends on the contractAI team (Odeist AI)From $29 a monthYes, in your voiceYes, on X and LinkedInYes, you approve or let it run The price of each tool is only part of the cost. Your time and your content count too, and we broke those down in [what social media automation really costs a small business](https://odeist.com/blog/social-media-automation-cost-small-business). Here is what each option is good and bad at. #### Doing it yourself Free, and nobody knows your product better. It works when one of you enjoys it and treats it like a standing meeting. A sample week for a founder selling, say, scheduling software for dental clinics: - Monday, 30 minutes: write three posts for the week in one sitting and schedule them. - Every weekday, 15 minutes: run five saved searches such as "any recommendations" dental scheduling or "tired of" no-shows, and reply to the three best recent posts. - Friday, 10 minutes: note which posts and replies brought profile visits or messages. The weakness is the week you ship. Product work always wins, and the routine lapses exactly when you have the most to say. Our guide to [finding leads in LinkedIn and X conversations](https://odeist.com/blog/find-leads-in-linkedin-and-x-conversations) has the search phrases that work best. #### A scheduler Buffer's free plan covers 3 channels with 10 scheduled posts per channel. Its Essentials plan is $5 a channel a month, billed yearly. Hootsuite's paid plans start at $99 a month, billed yearly. These figures are as of November 2026. A scheduler is the right buy if you already have things to say and your only problem is posting on time. It doesn't search for people asking about your problem, and it doesn't write replies. If posting is all you need, our guide to [scheduling posts to X, LinkedIn and Instagram at once](https://odeist.com/blog/schedule-posts-x-linkedin-instagram) covers how to plan the week. #### A freelancer On Upwork, the median rate for a social media manager is $20 an hour, with most between $14 and $35. Upwork's page for the United States gives $25 to $50. These figures are as of November 2026. The arithmetic looks friendly. Say you buy 10 hours a month at $20: that's $200. But look at where those hours go. Posts can be written from a brief. Good replies can't: they need someone who understands your product well enough to help a stranger with a specific problem, and who checks the searches every day, not every Tuesday. Ten hours a month is about 30 minutes a working day, which doesn't leave room for both. A freelancer is the right buy for work that rewards skill more than presence: a design system for your posts, a launch campaign, or a one-off content plan. #### An agency WebFX lists social media management from $3,150 a month, with a dedicated account manager, custom graphics and reporting. WebFX itself puts typical agency fees at $500 to $5,000 a month. Both figures are as of November 2026. For most two-person startups this is the wrong stage. Agencies earn their fee on strategy, design and paid ads at a scale you probably don't have yet. Come back to it when you have budget for ads and a brand to protect. #### An AI team This is the category Odeist AI is in, so here are the facts rather than the pitch. You paste your website, and Odeist AI writes your brand profile in about a minute: six documents (Product, Buyers, Competitors, Voice, Visual style and Action plan) that every agent works from. Then you turn on the agents you need: - Posting agents write and publish on X, LinkedIn and Instagram on a schedule you set. - Reply agents search X and LinkedIn for people asking for what you sell, asking about your competitors or describing your problem, and draft a reply in your voice. - A weekly review every Monday says what worked and what the team will change. Each agent has one switch. Ask me first sends every draft to you in WhatsApp, Telegram or Slack as a card with Approve, Edit and Skip. On its own lets that agent post by itself. Your skips and edits become rules the agent follows next time. What it doesn't do: replies run on X and LinkedIn only, not on Instagram. On Reddit, it works through Odeist Scout, our app on Reddit: Scout finds threads and drafts replies, and you post them yourself. Hacker News drafts and creator outreach always wait for you. ### What does a month cost with an AI team? Odeist AI charges credits per job, so the bill follows the work. A text post is 5 credits, a reply 2, a post with an AI image 30 and an article with its image 55. Reply agents also use a little for each search they run. A sample month for one brand: 20 text posts, 100 replies and 4 articles. At those prices it fits inside Solo's 3,500 credits, which cost $29 a month. That's an example to show the arithmetic, not a promise of results. The [pricing page](https://odeist.com/pricing) prices a busy month line by line. One detail matters for a team of two. Solo includes 1 seat. If both founders want to approve drafts, Growth includes 3 seats and 10,000 credits for $79 a month. Every paid plan has every agent; they differ only in credits, brands and seats. ### How do you pick the right alternative? Start from the job you're missing: If this is your problemPickYou have things to say but post irregularlyA schedulerYour posts look amateur, or you have a launch comingA freelancer, for a fixed projectYou have an ad budget and need a strategy around itAn agency, laterYou need customers, and nobody has time to find the conversationsAn AI team, or the do-it-yourself routineYou have no budget at allThe do-it-yourself routine Then test it before you commit: - Pick one network, the one where your buyers already talk. Founders and developers tend to be on X; business buyers on LinkedIn. - Run it for two weeks with the same effort every day. - Count three things: profile visits, replies to your replies, and private messages. - Keep it or switch. If a channel brings conversations, add the second network. If it doesn't, change your searches before you change the tool. Whatever you choose, keep a few things in your own hands: complaints, sensitive messages and big announcements. We wrote about where to draw that line in what an AI marketing agent should and shouldn't do without you. ### The cheap alternative that does the finding and the replying If your missing jobs are finding the conversations and answering them, that's what Odeist AI is built for. It reads your website, learns your voice, finds people on X and LinkedIn asking for what you sell, drafts replies and posts, and sends each one to your phone for a one-tap yes until you trust it to run on its own. Two of our customers put it this way: "When Chivvy was two people doing everything, Odeist AI was our social media team. It found people asking for what we’d built, and the Odeist AI team treated our launch like their own." — Chivvy "Odeist AI’s scheduled posts keep Nuack’s Instagram, X and LinkedIn alive, and we approve everything from WhatsApp in seconds. It’s like having a social media manager on the team." — Nuack You can try the Growth plan free for 7 days with 1,000 credits, and you don't need a card. Plans start at $29 a month after that; see [pricing](https://odeist.com/pricing). [Start free](https://odeist.com/signup) and turn on one reply agent first. Not ready to sign up? [Get a free growth report](https://odeist.com/audit): enter your website and see what your marketing is missing, free, in about a minute. ### What is the cheapest alternative to a social media manager? Doing it yourself costs nothing but your time, and a scheduler adds on-time posting for a few dollars a month. Neither finds people to reply to. An AI team like Odeist AI posts and drafts replies from $29 a month. ### Is a freelancer cheaper than an agency for a small startup? Usually, yes. On Upwork the median social media manager charges $20 an hour, while WebFX's agency plans start at $3,150 a month (both as of November 2026). A freelancer works best on a fixed project, like a launch, rather than daily replies. ### Can an AI tool replace a social media manager? It can take over the daily work: writing posts, publishing them, finding conversations on X and LinkedIn and drafting replies in your voice. With "Ask me first" on, nothing goes out until you approve it. Complaints, sensitive messages and big announcements are still best handled by you. ### Which Odeist AI plan suits two founders? Solo is $29 a month with 3,500 credits and 1 seat. If both founders want to approve drafts, Growth is $79 a month with 10,000 credits and 3 seats. ### Can I try Odeist AI before paying? Yes. New accounts get 7 days of Growth with 1,000 credits, and you don't need a card. ## AI Reply Approval: When One Tap Beats Autopilot, and When It Doesn't Source: https://odeist.com/blog/ai-reply-approval (updated 2026-10-20) Short answer: yes, approve AI replies to strangers before they post. A reply lands under someone else's post, with your name on it, and X's own automation rules say automated replies to posts found by keyword searches alone are not permitted. Full autopilot makes sense for scheduled posts on your own account, once you've stopped editing them. Keep the review cheap: a card that shows their post and your draft, three buttons, and two or three short check-ins a day. In short - Replies are riskier than posts: a stranger's thread, your name, a screenshot that lasts. - On X and LinkedIn, the platform rules lean towards a person deciding who gets a reply. - A good approval card shows the post you're answering, not only your draft. - Check four things per card: right person, true facts, your voice, no pitch. - Three check-ins of a few minutes each beat one big weekly clear-out. ### Should you approve AI replies before they post? For replies to people who haven't talked to you yet, yes. Not because AI drafts are bad, but because a reply is a different kind of message from a post. A post on your accountA reply to a strangerWho sees itYour followersTheir followers, and everyone reading that threadWhose spaceYoursTheirsContext the AI needsYour brandYour brand, and what this one person actually meantCost of a missA weak post you deleteA first impression you can't take backPlatform rulesMostly about spamSpecific rules on automated replies The AI can do the slow parts on its own: watching searches, picking threads, writing the draft. The one step worth keeping for yourself is the moment it goes out. If you want the general version of this rule for every kind of marketing work, we covered it in what an AI marketing agent should and shouldn't do without you. This article is only about replies. ### What do X and LinkedIn actually allow? Read the rules yourself before you automate anything. Here is what they say as of October 2026. X. The [automation rules](https://help.x.com/en/rules-and-policies/x-automation) (updated April 2026) say, about replies and mentions: "Automating these actions to reach many users on an unsolicited basis is an abuse of the feature, and is not permitted. For example, sending automated replies to posts based on keyword searches alone is not permitted." Automated replies are allowed when the person opted in first, for example by replying to your post or sending you a Direct Message. The same page reminds account owners: "You are ultimately responsible for the actions taken with your account, or by applications associated with your account", and that X may respond by "filtering your posts from search results or suspending your account." LinkedIn. The [User Agreement](https://www.linkedin.com/legal/user-agreement) (section 8.2, the Don'ts) says not to "Use bots or other unauthorized automated methods to access the Services, add or download contacts, send or redirect messages, create, comment on, like, share, or re-share posts, or otherwise drive inauthentic engagement". Our reading, which is not legal advice: a reply you have read and chosen to send is your decision to answer that person, not a keyword trigger firing by itself. That is the practical reason to keep a human tap on every reply to a stranger, on both networks. ### When does one-tap approval beat autopilot, and when doesn't it? Approval wins whenever a reply reaches someone new or says something that could be wrong. Autopilot wins when the work stays on your own account and the drafts have stopped surprising you. The jobOur pickWhyReplies to strangers found by keyword search on XAsk me firstX's rule on keyword-based automated repliesComments on strangers' LinkedIn postsAsk me firstYour name under their post; LinkedIn's rule on unauthorized automated commentsAny reply that names a price, a competitor, a customer or a numberAsk me firstWrong facts are the costliest mistakeThreads about layoffs, health, money trouble or breaking newsSkip, or Ask me firstA cheerful product reply reads badly thereScheduled posts on your own account, after weeks of approving them unchangedOn its own can fitYour space, your audience, easy to delete Approval also has failure modes. Be honest about these three: - You approve without reading. That is autopilot with extra taps, and worse, because you feel covered. - You check once a day. A reply is worth most while the thread is fresh. If drafts wait until evening, either add a midday check-in or narrow the agent to fewer, better threads. - The queue is so long you resent it. The fix is better searches and fewer drafts, not switching approval off. Our guide to [finding leads in LinkedIn and X conversations](https://odeist.com/blog/find-leads-in-linkedin-and-x-conversations) covers how to tighten the searches. ### What should a reply approval card show? You should be able to decide from the card alone. If you have to open X to understand it, the card is incomplete. - Their post, word for word, with who wrote it and when. You can't judge a reply without the question. - How busy the thread is. Two replies and two hundred are different decisions. - Why it was picked: which search or keyword matched. - Your draft, exactly as it will post. - Which account and brand it goes out from, if you run more than one. - When it posts: now, or a set time. - Approve, Edit, Skip. And when you skip, a quick reason. The WhatsApp side of this (cards, editing in plain words, the 24-hour rule) is covered in [how to approve social posts from WhatsApp](https://odeist.com/blog/approve-social-posts-from-whatsapp). ### What are you checking for in ten seconds? Read their post first, then the draft. Four questions: - Right person? Do they sound like a buyer, or a competitor, a student, a bot? - True? Every number, feature and price in the draft must be real. AI drafts sometimes fill gaps with plausible guesses. - Your voice? Would you have written it, or does it read like a press release? - Helpful first? Does it answer them before it mentions you? Say you sell scheduling software for physiotherapy clinics. The post: "Our clinic still books everything by phone. Is there a scheduler that handles no-shows without being a nightmare to set up?" Posted 40 minutes ago, three replies. Draft to skipDraft to approve"We built exactly this. ClinicFlow cuts no-shows by 60% and sets up in 5 minutes. Try it free: [link]""Two things fix most no-shows: a text the day before with a one-tap reschedule, and a waitlist that fills cancelled slots. Both are hard with phone-only booking. Roughly how many appointments a week do you run?" The first draft fails three checks at once: a made-up statistic, a pitch before any help, and a link nobody asked for. The second answers the question and leaves the door open. If they reply, you have a conversation, and a follow-up from you is now welcome. For more on writing replies people answer, see [how to get customers from X replies](https://odeist.com/blog/get-customers-from-x-replies). ### A review routine that takes minutes a day This is a sample routine, not a rule. Adjust the times to when your buyers post. WhenTimeWhat you doMorning, with coffeeAbout 5 minutesClear the overnight cards. Under 30 seconds each.LunchAbout 3 minutesCatch the threads that started this morning, while they are fresh.End of dayAbout 3 minutesClear what's left. Skip anything you're unsure about.FridayAbout 10 minutesRead your skip reasons. Drop a keyword that keeps finding the wrong people. Add one rule the drafts keep missing. Three habits make it fast: - One change per edit. "Shorter" or "drop the link", then look again. A list of five changes gets you a different draft, not a better one. - Skip with a reason, every time. "Too salesy", "wrong facts", "not our buyer". The reason is what improves tomorrow's drafts. - When unsure, skip. A missed thread costs nothing. A bad reply under your name costs more. This is where a tool should take the friction out. In Odeist AI, when you choose Not this one on a draft in Today, it asks what's off: Off-topic, Too salesy, Sounds like AI, Wrong facts, Wrong tone, or Something else. Your approvals, edits and skips become learned rules the agent follows on its next drafts, and you can edit or delete any rule. ### How do you know when to loosen the reins? Track one number for two weeks: how many replies you approve without changing a word. If almost every draft goes out untouched, the voice and the searches are right. That is the time to trim your check-ins, or to let scheduled posts on your own account run on their own. For replies to strangers, keep the tap. It is the cheapest insurance you have: a few seconds per reply against your name under the wrong post. ### How Odeist AI handles reply approval [Odeist AI](https://odeist.com/) is a team of AI agents for startups and small teams. Its X and LinkedIn agents find people asking for what you sell and draft replies in your voice, written from a brand profile it builds from your website in about a minute. - One switch per agent, under Before it goes out: "Ask me first" sends you every draft, "On its own" lets it post. Hacker News, Reddit and creator outreach always wait for you. - Approve where you are. Drafts wait in Today, and arrive as cards with Approve, Edit and Skip in WhatsApp, Telegram or Slack. Each card shows where it goes, the agent that wrote it, when it goes out and the words. - Edit in plain words. "Shorter" or "more casual" rewrites the draft; anything that posts or schedules asks for your yes. - Pay per reply, not per seat. A reply is 2 credits, plus a little for each search. Plans start at $29 a month; see [pricing](https://odeist.com/pricing). DataGait, one of our customers, put it this way: "Odeist AI became our outbound team on X and LinkedIn. It finds data engineers asking the questions DataGait answers and replies in our voice. The Odeist AI team tuned that voice with us on day one." Try it on the 7-day Growth trial with 1,000 credits, approvals included, no card needed. [Start free](https://odeist.com/signup) and watch the first drafts arrive. Or, if you'd like to see where you stand before you sign up: [Get a free growth report](https://odeist.com/audit). It shows what your marketing is missing, free, in about a minute. ### Should I approve AI replies before they post? For replies to people who haven't talked to you yet, yes. A reply lands in someone else's thread under your name, and X's automation rules say automated replies to posts found by keyword searches alone are not permitted. ### Is it against X's rules to auto-reply with AI? X's automation rules (updated April 2026) say sending automated replies to posts based on keyword searches alone is not permitted. Automated replies are allowed when the person opted in first, for example by replying to your post or sending you a Direct Message. ### When is it fine to let an AI agent post on its own? For scheduled posts on your own account, once you've approved them unchanged for a few weeks. Keep replies to strangers, and anything that names a price, a competitor or a number, on approval. ### How long does reviewing AI replies take each day? With a card that shows their post and your draft, most decisions take under 30 seconds. A sample routine is three check-ins of a few minutes each, plus about 10 minutes on Friday to read your skip reasons. ### Can I approve replies from my phone with Odeist AI? Yes. Set the agent to "Ask me first" and each draft arrives as a card with Approve, Edit and Skip in WhatsApp, Telegram or Slack, as well as in Today. You can try it free for 7 days, no card needed. ## How to Find Your First 100 Users on X and LinkedIn: A 30-Day Plan Source: https://odeist.com/blog/find-first-users-x-linkedin (updated 2026-10-14) Short answer: your first 100 users on X and LinkedIn come from three places, in this order: people who already know you, strangers posting about your problem right now, and people who watch you build. Spend days 1 to 3 messaging your network one person at a time, weeks 1 and 2 on a daily 30-minute search-and-reply routine, week 3 on posts made from what users tell you, and week 4 on cutting whatever brought nobody. Log every conversation in one sheet, so you know what actually worked. In short - A launch post is a bonus, not the plan. Message people by name first. - 100 users is a number of conversations, not a number of posts. - Search for the problem, reply in public, and message privately only after they answer. - Set up your first users yourself, on a call if you can. - Every Friday, count users by source. Keep what worked, drop the rest. ### Where do your first 100 users on X and LinkedIn come from? A launch post reaches your followers, and on a young account those are mostly friends and other founders. Users come from three pools, and each needs a different move. PoolWho they areHow you reach themWhenYour networkFormer colleagues, LinkedIn connections and X mutuals who fit your buyer, or know oneA personal message, one at a timeDays 1–3People asking nowStrangers describing your problem, comparing tools or complaining about a competitorA useful public reply, then a private message once they answerEvery day from week 1People watching you buildAnyone who sees your replies and clicks your profilePosts made from real user conversationsFrom week 3 Work your target backwards. Say you sell a tool that writes CRM notes from sales calls, and you guess that one in four people who talk to you will try it. Then 100 users means about 400 real conversations, or roughly 15 a day for a month. Your ratio will be different. The point stands: you need a daily conversation habit, not a better launch post. ### What should you set up before day 1? One hour, the day before you launch: - Fix both profiles. One line on who you help and what changes for them, a real photo, and a link that goes straight to sign-up. Our [playbook for getting customers from X replies](https://odeist.com/blog/get-customers-from-x-replies) covers the profile in detail. - Pin one post on X and add it to your LinkedIn Featured section. What it is, who it is for, one screenshot, how to try it. Everyone who clicks your name from a reply lands here. - Start a tracker. A spreadsheet with: name, network, source (network, a search phrase, or a post), first contact, replied, tried it, still using it after 7 days, and their exact words. - Write ten problem phrases the way a customer says them, not the way your landing page does. The [phrase table in our LinkedIn and X leads guide](https://odeist.com/blog/find-leads-in-linkedin-and-x-conversations) is a good start. - Decide your offer: "I'll set it up with you in 15 minutes." It gives a maybe a reason to say yes. ### Days 1–3: who should you message first? Not your feed. Your list. Go through your LinkedIn connections and X mutuals and write down 50 people who either have the problem or work next to someone who does. Then message each one by name. Weak launch messageStrong launch message"Hi, we just launched. Would mean a lot if you could check it out and share: [link]""Hi Priya, you once told me your team types up every sales call by hand. I just launched a tool that writes the CRM notes from the recording. Would you try it on two calls this week? I'll set it up with you, it takes 15 minutes." The strong one names a problem they have, asks for something small and specific, and offers your time instead of asking for theirs. For people who are not buyers, ask a different question: "Who is the one person you know who still does this by hand?" An introduction from someone they trust is usually the warmest lead you will get all month. A sample three days: - Day 1: publish your launch post on X and LinkedIn, written separately for each. Send 20 personal messages. Reply to every comment. - Day 2: 20 more messages. Set up everyone who said yes, on a call if possible. Log each one. - Day 3: the last 10 messages, and one follow-up to anyone who said "later". Then start the search routine below. ### Weeks 1–2: how do you find strangers who need it right now? Your network runs out in days. Strangers do not. Give this 30 minutes a day, at the same time: MinutesWhat you do0–10Run your saved searches: X on the Latest tab, LinkedIn filtered to Posts and sorted by latest.10–25Reply to the best five to ten posts. Answer the question; leave the link out.25–30Message anyone who replied to you yesterday. Update the tracker. In launch weeks, the best searches are about switching and timing, not just the problem. For the CRM-notes example: - "alternative to" [competitor] -filter:replies - "switching from" [competitor] - "is there a tool that" ("call notes" OR "sales calls") - "just hired our first" (sales OR SDR) - "how do you" "call notes" since:2026-11-01 On X, quotes match an exact phrase, OR matches either word, a minus sign excludes, and since: with last week's date keeps results recent. On LinkedIn, paste the same phrase, filter to Posts and read the comments too: the people writing "same problem here" are a second list. How to write the reply is covered in [the X replies playbook](https://odeist.com/blog/get-customers-from-x-replies): show you read the post, give one useful thing, end with a question. This routine is also the habit that tends to slip once early users start asking for fixes. That is the step [Odeist AI](https://odeist.com/) is built for: its X and LinkedIn agents run your searches on a schedule and draft a reply to each match in your voice, for you to approve from your phone. ### Week 2: how do you turn a reply into an active user? A reply starts a conversation. A user needs a second step. When someone answers you, move to a private message and offer to do the work for them. Weak follow-upStrong follow-up"Thanks, here's the sign-up link: [link]""Glad that helped. Notes going missing after back-to-back calls is exactly what I built this for. Want me to set it up before your next call? Ten minutes on a screen share." Paul Graham calls recruiting users by hand "the most common unscalable thing founders have to do at the start" in [Do Things That Don't Scale](https://paulgraham.com/ds.html). Setting people up yourself also gives you something no dashboard will: you watch where they get stuck. Seven days after each sign-up, check the tracker. Anyone who has not come back gets one question: "What stopped you?" Their answer is either a bug, a missing feature, or a sign they were never the buyer. All three are useful. ### Week 3: what should you post when nobody follows you yet? By now you have something most launch posts lack: real conversations. Post from them, three times a week, while replies stay your main work. PostExampleWhat you learned"I set up 20 sales teams' call notes by hand this month. Not one of them reads the notes they already have."A fix a user asked for"A user asked for notes in Spanish on Tuesday. It shipped Thursday." Credit them if they agree.How people cope todayThe spreadsheet, the voice memo, the intern. Show the workaround and what it costs them.A plain how-toAnswer a question you saw three times in your searches this week. Write each idea once and adapt it for each network, which we cover in [how to schedule posts to X, LinkedIn and Instagram at once](https://odeist.com/blog/schedule-posts-x-linkedin-instagram). Use your users' words, never their names, unless they say yes. ### Week 4: what should you double down on? Every Friday, sort the tracker by source and answer three questions: - Which sources brought people who are still using it? Not sign-ups. Users still there after a week. - Which search phrases brought nothing? Delete them. Replace them with phrases your users actually said on calls. - Who are your five happiest users? Ask each one: "Who else do you know with this problem?" and whether you can quote a sentence of theirs. Then plan month two around the top two sources. Expect that to mean fewer, better searches and more time in the conversations they produce. ### Which mistakes stall the first 100? - Launching to a feed instead of to people. Likes on a launch post are not users. - Searching for your category name. People describe their problem, not your product. - Not logging the source. Without it, week 4 is guesswork. - Stopping the routine when you get busy. The week you ship fixes is the week the pipeline empties. ### How can Odeist AI run this playbook with you? Odeist AI is a team of AI agents for startups and small teams. For the first 100 users, it takes on the parts that slip: - It learns your business in about a minute. Paste your website and it writes six documents every agent works from: Product, Buyers, Competitors, Voice, Visual style and an Action plan, week by week. It can also learn your voice from your own posts. - It finds the conversations. Your X and LinkedIn agents search for the keywords you give them, such as your problem phrases, and draft a reply to each match in your voice. - You keep the final say. Each agent has one switch. "Ask me first" sends every draft to WhatsApp, Telegram or Slack with Approve, Edit and Skip. Once the drafts sound like you, switch to "On its own" and it posts by itself. Skips and edits become rules the agent follows. Our guide to what an AI marketing agent should do without you explains how to choose. - It helps with launch day. The Launch kit writes an X thread, a LinkedIn post, a launch email, a press release, three taglines and a plan from a week before to a week after. The Hacker News agent drafts your Show HN, which you post yourself. - It reports back. Every Monday the team reviews the week and says what it will change. A reply uses 2 credits, and plans start at $29 a month; see [pricing](https://odeist.com/pricing). New accounts get 7 days of Growth with 1,000 credits, and no card. "When Chivvy was two people doing everything, Odeist AI was our social media team. It found people asking for what we’d built, and the Odeist AI team treated our launch like their own." — Chivvy [Start free](https://odeist.com/signup), or [get a free growth report](https://odeist.com/audit) to see what your marketing is missing, in about a minute. ### How do I find my first users on X and LinkedIn after launching? Message people in your network who fit your buyer one at a time, then spend 30 minutes a day searching both networks for people describing the problem you solve and replying helpfully. Move to a private message only after they answer, and offer to set the product up with them. ### How long does it take to get the first 100 users? It depends on your product and your ratio of conversations to sign-ups. Work it backwards: if one in four people you talk to tries it, 100 users means about 400 conversations, so plan a daily habit for at least a month. ### Should I put the link in my replies? No. Answer the question in the reply and let your profile do the selling, with a pinned post and a sign-up link. Share the link in a private message once someone asks or says yes. ### What should I post on X and LinkedIn when I have no followers yet? Post from your user conversations: what you learned, fixes users asked for, how people cope with the problem today, and plain how-tos for questions you keep seeing. Replies stay your main work until you have an audience. ### Can an AI agent find and reply to these posts for me? Yes. Odeist AI’s X and LinkedIn agents search for your keywords and draft replies in your voice; with “Ask me first” every draft waits for your approval in WhatsApp, Telegram or Slack. A reply uses 2 credits, and the 7-day Growth trial needs no card. ## How to Find Customers on X and LinkedIn: The 5-Step B2B Playbook Source: https://odeist.com/blog/find-customers-on-x-and-linkedin (updated 2026-10-08) Short answer: to find customers on X and LinkedIn, write down the questions your buyers ask before they buy, turn each one into a saved search on both networks, and reply to the fresh posts from real buyers with a useful answer, not a pitch. Do it on a fixed daily routine, have one person approve what goes out, and prune your searches every week. That is how B2B teams like DataGait reach their buyers: by answering the questions those buyers already ask in public. In short - Start from your buyers' questions, sorted by how close each one is to buying. - Build searches from their words, with operators that cut the noise. - Score each post before you reply. Most are not worth it. - Match the reply to the question. Mention your product only when it fits, and say you built it. - Run it daily, approve before it posts, and drop searches that find nobody. ### How do B2B teams like DataGait find customers on X and LinkedIn? DataGait sells to data engineers. Here is how the team describes its approach, in its own words: "Odeist AI became our outbound team on X and LinkedIn. It finds data engineers asking the questions DataGait answers and replies in our voice. The Odeist AI team tuned that voice with us on day one." That one sentence holds the whole method. A clear buyer (data engineers). A trigger (a question the product answers). An action (a reply that sounds like the team). No bought list, no cold messages to strangers. The Applier uses the same pattern for a different buyer: "Odeist AI puts The Applier in the LinkedIn and X threads where job seekers ask for help. More new users now find us through those replies." The rest of this guide is that loop, step by step. The searches and replies below are our own examples for a made-up company that sells a data pipeline monitoring tool to data engineers. They are not DataGait's. ### Step 1: Which questions do your buyers ask before they buy? Most teams start with keywords. Start with questions instead, because that is what people post. Mine them from places you already have: - Sales and demo calls. Ask every new customer: "What were you trying to fix when you started looking?" - Support tickets and onboarding emails. The first question a new user asks is often the one they posted publicly a week earlier. - Competitor names. People post when they are leaving a tool. - Your own FAQ page. Rewrite each question the way a tired engineer would type it. Aim for 15 to 20 questions, then sort them by how close the person is to buying: StageExample post (data pipeline tool)What your reply should doPain"Third time this month a schema change silently broke our dashboards."Diagnose. Give one fix they can try today.Method"How are you all testing dbt models before they hit prod?"Share how teams handle it, with a concrete detail.Shortlist"Any recommendations for data quality monitoring that isn't enterprise-priced?"Name honest options, yours included, and say who each suits.Switching"Looking for an alternative to [tool], pricing just doubled."Compare fairly. Offer a tip that makes the move easier. Shortlist and switching posts are the warmest. Pain and method posts are the most common, and they are where people first see your name. ### Step 2: How do you turn those questions into searches on X and LinkedIn? Each search is a phrase from the question, a word from your field, and a few words to exclude. The two networks read searches differently. On X, wrap phrases in quotes, use OR for alternatives and a minus sign to exclude a word. -filter:links drops posts with links, which are often promotion, and lang:en keeps one language. After searching, switch to the Latest tab. - "schema change" (broke OR broken) dbt -filter:links lang:en - "how do you" testing "dbt models" -hiring - "alternative to" ("data quality" OR observability) -filter:links On LinkedIn, the search bar takes Boolean strings: quotes for exact phrases, AND, OR and NOT typed in capitals, and parentheses for grouping. LinkedIn's help page says minus signs and wildcards aren't officially supported, so use NOT. Then pick Posts, sort by Latest and narrow Date posted to the past week. - "data quality" AND ("any recommendations" OR "what do you use") - ("schema change" OR "schema drift") AND (broke OR broken) NOT hiring XLinkedInExclude a word-hiringNOT hiringOperatorsQuotes, OR, minusQuotes, AND / OR / NOT in capitals, parenthesesNewest firstThe Latest tabSort by Latest, Date postedExtra filters-filter:links, lang:enPosts only Then test each search. Run it and count how many of the first 20 results are a real buyer asking a real question. Keep the searches that find at least three. Rewrite or drop the rest. You should end up with eight to twelve searches worth running every day. ### Step 3: Which posts deserve a reply today? A good search still returns more posts than you should answer. Score each one from 0 to 2 on five things before you write a word: Check02FitStudent, vendor or recruiterYour buyer, by role and companyStageSharing newsShortlisting or switchingFreshnessDays old on XPosted in the last few hoursRoomDozens of replies alreadyFew or noneSafetyA rant, a pile-on or politicsA calm, open question Reply to anything at 7 or more. A LinkedIn post can still score on freshness after a day or two, since conversations there move slower. For more on reading a post's author and fit, see [how to find leads in LinkedIn and X conversations](https://odeist.com/blog/find-leads-in-linkedin-and-x-conversations). ### Step 4: What should a reply to a buyer actually say? The basics of a good reply (show you read the post, give one useful thing, end with a question) are in [our playbook for getting customers from X replies](https://odeist.com/blog/get-customers-from-x-replies). What changes in B2B is the stage. A pain post wants a diagnosis. A shortlist post wants honest options. Pitching to someone who only asked how others do it is how a reply gets ignored. Here is an example for the method question above: BeforeAfter"You should check out [product], it solves exactly this. Link in bio.""We put a contract test on the columns our dashboards read and fail the run on a rename or a type change, not just a spike in nulls. It catches most breaks before prod. Are yours mostly app changes upstream or vendor feeds?" If they answer and ask what you use, that is the moment to mention your product, and to say you made it: "Full disclosure, I work on [product], which runs that check for you. Happy to share how to set it up by hand too." On X, keep it to two or three lines. On LinkedIn, a few short sentences and one question read naturally. On both, write like an engineer talking to an engineer: plain words, one specific detail, no adjectives doing the selling. ### Step 5: How often should you search and reply? Threads go cold, so this works as a routine, not a campaign. Here is a sample week for one person: WhenWhatTimeMondayReview last week. Drop searches that found nobody. Add two questions from recent sales calls.30 minEvery weekday, twiceRun the searches, score the posts, reply to the best five to ten. Answer anyone who replied to you.15 minFridayCount replies sent, replies that got an answer, conversations moved to messages, and sign-ups or demos that mention a thread.15 min Judge the channel on conversations and sign-ups, not likes, and give it a month before you decide. ### Who should approve replies before they go out? Every reply speaks for your company in public, so one person should own the yes. A simple rule set: - Start by approving every reply. You will learn fast which drafts are safe. - Always keep these for a human: complaints about your own product, threads about pricing or security, and angry posts about a competitor. - Hand over the rest gradually, once the drafts stop needing edits. More on drawing that line in what an AI marketing agent should and shouldn't do without you. This is the step where most teams stall, and where Odeist AI fits. Its X and LinkedIn agents run your keywords on a schedule, draft each reply in your voice, and send each draft to the app, WhatsApp, Telegram or Slack as a card with Approve, Edit and Skip. See [how approving from WhatsApp works](https://odeist.com/blog/approve-social-posts-from-whatsapp). ### What makes this stop working? - Searching for your category instead of the question. Nobody posts "seeking data observability platform". - A search list that never changes. Buyers' words change. Your list should too. - Replying to everything. Ten good replies beat forty thin ones. - Counting likes. A reply with no likes that starts a conversation with a buyer is the win. - Hiding who you are. If you work on the product, say so when you mention it. ### How does Odeist AI run this playbook for you? Odeist AI is a team of AI agents that finds the people looking for what you sell, answers them in your voice and reports back. For this playbook, it takes over the searching, the drafting and the daily routine. - It learns your buyers and voice. Paste your website and it writes six brand documents, including Buyers and Voice, in about a minute. It can also learn how you write from your own posts. - It finds the questions. The X and LinkedIn agents look for people asking for what you sell or about your competitors. You set the keywords, your rules on who to answer, and how often they look. - It drafts in your voice. Each reply starts from your brand documents. - You choose the switch. Each agent has "Ask me first" or "On its own". Every skip and edit teaches the agent a rule it follows next time. - It costs little per reply. A reply is 2 credits, plus a little for each search. Every plan gets every agent, from $29 a month. See [pricing](https://odeist.com/pricing). Chivvy put it this way: "When Chivvy was two people doing everything, Odeist AI was our social media team. It found people asking for what we’d built." And the wrkr: "We’re pre-launch and Odeist AI is already growing our waitlist, joining conversations about flaky end-to-end tests." Try it with 7 days of Growth and 1,000 credits, no card needed. [Start free](https://odeist.com/signup), or enter your website and [Get a free growth report](https://odeist.com/audit) to see what your marketing is missing. Whether you use a tool or not, the loop is the same: know the questions, search for them every day, and be the most useful reply in the thread. ### How do I find customers on X and LinkedIn? List the questions your buyers ask before they buy, turn each into a saved search on both networks, and reply to fresh posts from real buyers with a useful answer. Run the searches every weekday and drop the ones that find nobody. ### What should I search for on X and LinkedIn to find buyers? Search for a phrase from your buyer's question plus a word from your field, such as "alternative to" or "any recommendations", and exclude noise like hiring posts. On X use quotes, OR and a minus sign; on LinkedIn use quotes and AND, OR and NOT in capitals, then sort posts by Latest. ### Should I message people directly or reply in public first? Reply in public first, with an answer to their question. Move to private messages only once they reply, and mention your product only when it fits, saying that you work on it. ### How much time does this take each week? About 15 minutes twice a weekday to run searches and reply, plus 30 minutes on Monday to prune searches and 15 minutes on Friday to count conversations and sign-ups. Give it a month before judging the channel. ### Can a tool do this without posting replies I haven't seen? Yes. In Odeist AI, each X or LinkedIn agent has a switch: "Ask me first" sends every draft to you in the app, WhatsApp, Telegram or Slack, and "On its own" posts by itself. A reply is 2 credits, and the 7-day Growth trial needs no card. ## How to Find Leads in LinkedIn and X Conversations (Without Scrolling All Day) Source: https://odeist.com/blog/find-leads-in-linkedin-and-x-conversations (updated 2026-10-24) Short answer: to find leads on LinkedIn and X, search for the words people use when they have the problem you solve ("any recommendations for", "struggling with", "looking for a tool"), reply helpfully to the recent ones, and move to private messages only once they reply. Save your best searches and check them daily, or let a tool watch them for you. In short - Leads show up as questions, complaints and comparisons, not as ads. - Search for problem phrases, not your product category. - Reply in public first. Message privately only after they respond. - Check your saved searches once a day, not all day. ### What does a lead look like on social media? Most people never write "I want to buy a product like yours". They write things like: - "Does anyone know a good way to…" - "We've tried X and Y and neither works for us." - "How are other teams handling…?" - "Honestly so tired of doing this by hand." Each of those is a person telling you, in public, that they have a problem and are open to solutions. The job is finding them before someone else does. ### What should you search for? Combine a problem phrase with a topic word from your industry. Problem phraseWhat it signals"any recommendations for"Actively looking. The warmest lead."alternative to"Unhappy with what they use now."struggling with" / "tired of"Feeling the pain, may not be looking yet."how do you" / "how are you"Asking peers, open to ideas."we just hired" / "we just launched"New situation, new needs. For example, a payroll service might search "any recommendations" payroll or "tired of" running payroll. ### Where to search on each platform On X: use the search bar and switch to the Latest tab so you see new posts first. Wrapping a phrase in quotes finds that exact phrase. On LinkedIn: search your phrase, filter to Posts, and sort by Latest. LinkedIn conversations move slower than X, so posts from the last few days are still worth replying to. Keep a short list of five to ten searches that consistently bring up good posts, and drop the ones that don't. ### Which posts are worth replying to? You'll find more posts than you have time for. Prioritise the ones where: - The person matches your customer. Check their profile: right role, right kind of company. - The post is recent. Fresh posts are still being read. - Few people have replied. You'll stand out more in a quiet thread. ### How do you turn a reply into a conversation? Your first reply should help, not sell. Answer the question, share what worked for you, or point out something they may have missed. If they respond, you've earned the right to continue. Then take it to private messages with something useful: a short checklist, a quick answer to their follow-up, or an offer to look at their situation. Mention what you do only when it genuinely fits. What to avoid: - Replying with a link and nothing else. - Sending a sales message to someone who hasn't replied to you. - Pasting the same reply under every post. ### How do you do this without scrolling all day? Give it a fixed 20 minutes a day: run your saved searches, reply to the best five to ten posts, and follow up with anyone who responded yesterday. If even that is hard to fit in, [Odeist AI](https://odeist.com/) can do the watching for you. It checks X and LinkedIn for posts that match your topic, writes replies in your brand's voice and posts them. You can also have every reply sent to your WhatsApp to approve before it goes out. Plans start at $29 a month, with a [7-day free trial](https://odeist.com/login), no card needed. ### What should I search for to find leads on LinkedIn and X? Search for the words people use when they have the problem, not the name of your product category. Phrases like “any recommendations for”, “struggling with” and “looking for a tool that” combined with your topic find people who are ready to talk. ### Is it OK to message someone after replying to their post? Yes, if the reply started a real conversation. Wait until they respond to you, then continue in private messages with something useful, not a sales pitch. ### How can I find leads on social media without spending hours a day? Save your best searches and check them once a day, or use a tool that watches those conversations for you and suggests replies. Odeist AI does this on X and LinkedIn, with plans from $29 a month. ## How to Schedule Posts to X, LinkedIn and Instagram at Once Source: https://odeist.com/blog/schedule-posts-x-linkedin-instagram (updated 2026-10-24) Short answer: to schedule posts to X, LinkedIn and Instagram at once, plan your ideas for the week in one sitting, write one version of each idea per platform, then load them all into a scheduling tool that posts to all three. The trick is not copying the same text everywhere. Keep the idea, change the shape. In short - Batch your planning: one sitting a week is enough. - Start from ideas, not from platforms. - Adapt each idea: short for X, a story for LinkedIn, visual-first for Instagram. - Post at a pace you can keep for months, not days. ### Step 1: How do you plan a week of posts in one sitting? Block 45 minutes once a week. Write down five ideas worth sharing. Good sources: - A question a customer asked you this week. - A mistake you see people in your industry make. - A small result: a happy client, a number that went up, a lesson learned. - Something behind the scenes: how you make, pack or deliver what you sell. - Your opinion on news in your field. Five ideas are enough for a full week across all three platforms. ### Step 2: How should one idea look on each platform? Each platform rewards something different. Here's how one idea, "most small businesses post too rarely, not too little", changes shape: PlatformWhat worksExampleXShort, direct, one clear point"Small businesses don’t fail on social because their posts are bad. They fail because they post twice and stop."LinkedInA short story or lesson, with space between lines"Last year we posted 4 times in January and then nothing until May. Here’s what changed when we switched to 3 posts a week…"InstagramA strong image or short video first, short caption secondA calendar photo with the caption "3 posts a week, every week. That’s the whole strategy." ### Step 3: How often should you post on each platform? A steady pace you can keep beats a burst you can't. A sensible starting point for a small business: - X: once a day - LinkedIn: three times a week - Instagram: three times a week That's 13 posts a week from five ideas. Some ideas become two or three posts. Others only suit one platform, and that's fine. ### Step 4: What's the best time to schedule posts? There's no single best time for everyone. Start with when your customers are likely to be on their phones: early morning, lunchtime and early evening in their time zone. After two or three weeks, look at which posts got the most replies and move more of your schedule to those times. ### Step 5: How do you schedule everything at once? You need a tool that connects to all three accounts and publishes for you. When comparing options, check: - Does it post to all three natively? Some tools only send you a reminder for Instagram instead of posting. - Can you see everything on one calendar? It makes gaps obvious. - Can it help write the platform versions? This is where most of your time goes. - What does it cost per account? Many tools charge for features you'll never use. [Odeist AI](https://odeist.com/) connects to X, LinkedIn and Instagram, writes posts in your brand's voice, and publishes them on a calendar you can see at a glance. It can also make short AI videos for your posts. Plans start at $29 a month, and every paid plan posts to all three. You can [try it free for 7 days](https://odeist.com/login), no card needed. ### Common mistakes to avoid - Copying and pasting the same text everywhere. It shows, and each audience notices. - Scheduling a month ahead and never checking. Look in once a day and reply to comments. Replies are where conversations, and customers, come from. - Posting only about your product. Aim for roughly one promotional post in every five. The rest should be useful or interesting on their own. ### Can I post the same thing on X, LinkedIn and Instagram? You can share the same idea, but adjust the wording for each. X rewards short, punchy posts. LinkedIn rewards a clear story or lesson. Instagram needs a strong image or video first and a short caption second. ### How often should a small business post on each platform? A steady pace you can keep beats a burst you cannot. A good starting point is once a day on X, three times a week on LinkedIn and three times a week on Instagram. ### What is the cheapest way to schedule posts on all three platforms? With Odeist AI, plans start at $29 a month, and every paid plan posts to X, LinkedIn and Instagram. You can try it free for 7 days first. ## How Much Does Social Media Automation Cost for a Small Business? Source: https://odeist.com/blog/social-media-automation-cost-small-business (updated 2026-10-24) Short answer: a small business can automate its social media for $29 to $79 a month. The real cost has three parts: the tool, the time you still spend, and the content you feed it. For most small teams the tool is the smallest of the three, and the time it saves is the biggest. In short - You pay for three things: a tool, your time and your content. - Most small businesses only need one or two social accounts to start. - Count the hours automation saves you, not just the monthly fee. - Start with the one channel where your customers already are. ### What are you actually paying for? When people ask what social media automation costs, they usually mean the monthly fee. That's only part of it. CostWhat it coversTypical rangeThe toolScheduling posts, finding conversations, writing drafts$29 to $79 a month for a small businessYour timeChecking what goes out, answering people who reply15 to 30 minutes a dayContentImages, videos, product photos, your own ideasFrom free (your phone) to paid design help ### How much does your time cost without automation? This is the number most people forget. Say you spend 45 minutes a day finding things to post, writing captions and replying to comments. That's about 5 hours a week, or 20 hours a month. If your time is worth $30 an hour, that's $600 a month of your time. Cut it to 15 minutes a day with automation and you get back roughly 15 hours, or $450 a month, for a tool that costs a fraction of that. Swap in your own hourly rate. The maths is almost always in favour of automating the repetitive parts. ### What does it cost with Odeist AI? Odeist AI sells plans in credits. Every paid plan has every agent and approvals on WhatsApp; the plans differ only in credits, brands and seats: - Solo: $29 a month for 3,500 credits, with 1 brand and 1 seat. - Growth: $79 a month for 10,000 credits, with 3 brands and 3 seats. - Custom: for agencies and bigger teams, any amount a month at about 160 credits a dollar ($200 a month for about 32,000 credits), with the brands and seats you need. Talk to us. - Free: 250 credits to start, once, to try the agents on 1 brand. Paying yearly gives you 2 months free. Each job has a price in credits, worked out from what it costs to run: JobCreditsA post, text only5A post with an AI image30A reply, drafted in your voice2An article with its image55An 8-second video clip with sound704 The [pricing page](https://odeist.com/pricing) prices a busy month for one brand line by line and shows which plan it fits. ### Which channel should a small business automate first? Start where your customers already spend time, not where you think you should be: - Selling to other businesses? Start with LinkedIn. - Selling to founders, developers or a tech crowd? Start with X. - Selling something visual, or to local customers? Start with Instagram. Get one channel working before adding a second. One account run well beats three accounts run badly, and it keeps you on the smallest plan while you learn what works. ### What should you never automate? Automation is great for the repetitive work. Some things still need you: - Complaints and sensitive messages. Answer these yourself. - Big announcements. Write and post these by hand. - Anything you haven't checked. If you're starting out, have every post and reply sent to your WhatsApp to approve before it goes live. ### How to start without wasting money - Pick one channel using the list above. - Try it free first. Odeist AI has a [7-day free trial](https://odeist.com/login), and you don't need a card. - Check results after the trial: new followers, profile visits and conversations. - Only then add a second channel or move to a bigger plan. You can also [see what each job costs](https://odeist.com/pricing) before you sign up. ### What is the cheapest way to automate social media for a small business? Start with the one or two channels that bring you customers. With Odeist AI, plans start at $29 a month, and there is a 7-day free trial with no card needed. ### Is social media automation worth it for a small business? It is worth it when it replaces hours of your own time each week. If you value your time at $30 an hour and automation saves you three hours a week, that is roughly $360 a month of time saved, compared with a tool that costs $29 to $79. ### How much does Odeist AI cost? Solo is $29 a month for 3,500 credits and Growth is $79 for 10,000; agencies get a Custom plan ($200 a month for about 32,000 credits). Every paid plan has every agent and approvals on WhatsApp. A text post is 5 credits and a post with an AI image 30. There is a 7-day free trial, no card needed. ### Do I need a social media manager if I use automation? Not for day-to-day posting and replying. Many small businesses use automation for the daily work and spend 15 minutes a day checking what goes out and answering people who reply. ## How to Get Customers From X (Twitter) Replies: A Simple Playbook Source: https://odeist.com/blog/get-customers-from-x-replies (updated 2026-10-24) Short answer: to get customers from X, spend less time posting and more time replying. Find posts from people who already have the problem you solve, add something genuinely useful underneath, and let them click through to your profile. Ten to twenty good replies a day, every day, will do more for a small business than one clever post a week. In short - Replies put you in front of people who are already talking about your topic. - Search for the problem, not your product name. - Useful beats clever. Answer the question fully, and skip the link. - Make your profile do the selling: one clear line about who you help. - Be consistent. A little every day beats a lot once a week. ### Why do replies work better than posts on X? When you post, only your followers see it, and a new account has very few. When you reply, you show up in front of everyone reading that conversation. The person who wrote it already has an audience, and people reading a thread about a problem are, by definition, interested in that problem. That makes replies the cheapest way to borrow attention you haven't earned yet. You don't need a big following. You need to be in the right conversations. ### Which posts should you reply to? Not every post is worth your time. The best ones share three things: - The writer has the problem you solve. They're asking for help, comparing options, or venting about something your product fixes. - It's recent. Most people read a post while it's fresh, so an early reply tends to be seen by far more of them than one a day later. - The conversation is still open. A question with two replies is a better bet than one with two hundred. To find them, search X for the words people use when they're stuck. If you sell bookkeeping software, don't search "bookkeeping software". Search things like: - "any recommendations" bookkeeping - "hate doing" invoices - "how do you" track expenses Save the searches that bring up good posts, and check them once or twice a day. ### How do you write a reply people actually answer? A good reply makes the reader's life a little easier, even if they never buy anything. A simple structure that works: - Show you read the post. Mention the specific thing they said. - Give one useful thing. A tip, a number, a short example from your own experience. - End openly. A light question invites them to respond. Here's the difference: Weak replyStrong reply"Great point! Check out our tool 👉 link""We had the same problem at 20 invoices a month. What fixed it was sending them the same day the work finished, not at month end. Are yours mostly one-off or repeat clients?""So true 🙌""The part about chasing late payments hits home. A polite reminder at 3 days and 10 days got most of ours paid without a phone call." Leave links out unless someone asks. Replies with links often look like ads and tend to get less attention. If your reply is helpful, the curious ones will click your name anyway. ### What should your profile say? Every good reply sends a few people to your profile. What they see there decides whether they follow, visit your site or leave. Keep it simple: - Name and photo that look like a real person or a clear brand. - One line on who you help and how. "I help freelancers get paid on time" beats a list of job titles. - One link to the page you want them to see. - A pinned post that shows your best idea or a customer result. ### How many replies should you send each day? Start with 10 to 20 replies a day. That's enough to be seen regularly in your niche without your replies turning generic. It takes about 20 to 30 minutes if you already have your searches saved. Track three things each week: how many people visited your profile, how many followed you, and how many conversations turned into private messages. Then do more of whatever brought those. ### Can you do this without spending all day on X? Yes. The slow part is finding the right posts, not writing the replies. That's the part worth handing off. [Odeist AI](https://odeist.com/) watches X and LinkedIn for posts that match your topic, suggests replies in your brand's voice and can post them for you. If you'd rather read everything first, Odeist AI can send each reply to your WhatsApp, and nothing goes out until you approve it. Plans start at $29 a month, and you can [try it free for 7 days](https://odeist.com/login) without a card. Whether you use a tool or not, the habit is what matters: show up in the right conversations every day and be the most helpful person in the thread. ### How many replies should I send on X each day? Start with 10 to 20 thoughtful replies a day. That is enough to be noticed in your niche without your replies starting to sound the same. Increase it once you know which kinds of posts lead to profile visits and conversations. ### Should I put a link to my product in my replies? Usually not. Replies with links often get less reach and look like ads. Answer the question fully, and let people click your profile. Share a link only when someone asks for one. ### Can I automate replies on X? You can use a tool to find the right posts and draft replies for you. Whatever you use, read the replies before they go out, keep them specific to the post, and follow X’s rules on automation. ### How long before replies bring in customers? It depends on your niche and how consistently you reply. Early signs are profile visits and new followers from people in your niche. Give it at least a few weeks of daily replies before judging whether it works. ## Careers at Odeist AI Odeist AI finds the people already asking for what a business sells, in live conversations on X and LinkedIn, and on Reddit through Odeist Scout, our app on Reddit, and replies in that brand's voice. Founders approve every reply and post from WhatsApp. We are a small, remote team that ships every week and talks to customers every day. ### Founding Full-Stack Engineer https://odeist.com/careers/founding-full-stack-engineer · Engineering · Full-time · Remote (India) Build the product founders use to find customers in social conversations: the agents that read X, LinkedIn and Reddit, the approval flow on WhatsApp, and the app that ties it together. You will be one of the first engineers. You will work across the React app, the Firebase Cloud Functions behind it, and the workers that talk to social platforms. Most weeks you will ship something a customer notices. ### AI Engineer, Agents https://odeist.com/careers/ai-engineer-agents · Engineering · Full-time · Remote (India) Make Odeist AI's agents better at finding buying-intent conversations, ranking them, and writing replies that sound like the brand, and prove it with evaluations. Our agents read thousands of posts a day and decide which few are worth a reply. You will own how well they do that: the ranking, the drafting, the voice, and the measurements that tell us a change made things better. ### Growth Marketer, Social and Community https://odeist.com/careers/growth-marketer · Marketing · Full-time · Remote (India) Grow Odeist AI on X, LinkedIn and Reddit using Odeist AI itself, and build the content, SEO and free tools that bring founders to us every day. Our product is social growth, so our own growth is the best demo. You will run the channels, write the guides and comparisons that rank, and own the funnel from first visit to paying customer. ### Product Designer https://odeist.com/careers/product-designer · Design · Full-time · Remote (India) Design how founders review, edit and approve what Odeist AI writes, in seconds and on a phone, and own the design system across the site and the app. Most of our customers use Odeist AI from WhatsApp, between meetings. You will make every approval feel quick and safe, and shape the new agents we are building before a line of code is written. ### Customer Success Lead https://odeist.com/careers/customer-success-lead · Customer Success · Full-time · Remote (India) Get new customers and agencies live in an afternoon, keep them growing, and bring what they tell us back into the product. Customers trust Odeist AI to speak for their brand. You will set them up so it sounds right from day one, and you will be the person they call when they want more. ## Contact - Website: https://odeist.com - Email: connect@odeist.com - X: https://x.com/odeist_ai - LinkedIn: https://www.linkedin.com/company/odeist-ai/