AEO Copilot to track what ChatGPT and Claude say about your brand

What happened

AEO Copilot Launches Free Brand-Tracking Tool for ChatGPT and Claude Mentions

Sofian Bettayeb, an SEO and AEO practitioner, has launched AEO Copilot, a web platform that monitors how ChatGPT, Claude, Perplexity, and Google AI Overview respond to questions about a brand. The free tier requires no credit card, carries no expiry, and covers one brand and 50 prompts per month tracked across ChatGPT and Claude. Paid plans starting at $9 per month extend coverage to all four engines and add API access, data export, and scheduling options up to weekly. The Freelancer plan at $34 per month and the Agency plan at $99 per month introduce white-label report branding and, on the Agency tier, unlimited brands and team access.

This article covers how AEO Copilot works mechanically, who it targets, what it does and does not do, and what builders evaluating it should verify before committing to a paid tier. It is written for freelancers managing client search visibility, marketing teams at startups navigating AI-native search, and developers who want to pipe brand-mention data into an agent-driven content workflow.

How it works

Sofian Bettayeb has released AEO Copilot at aeo-copilot.com, positioning it as an answer engine optimization tool for freelancers, agencies, and AI-native startups who cannot justify the pricing of enterprise-grade alternatives. The service asks users to supply the questions their customers actually ask — or lets onboarding draft the first nine — then runs those prompts across ChatGPT, Claude, Perplexity, and Google AI Overview on a user-configured schedule. Results come back as a dashboard report, a REST API response, or structured data fed to an AI agent via a standard Model Context Protocol server.

The free tier covers one brand and 50 prompts per month refreshed on the first of each month, requires no credit card, and has no trial countdown or expiry. The $9 Startup plan keeps the same prompt limit and brand count but adds all four engines, API and MCP access, data export, and the option to run checks twice monthly on the first and fifteenth or weekly every Sunday. The Freelancer plan at $34 expands to five brands and 250 prompts, and the $99 Agency plan allows unlimited brands with 600 prompts and role-based team access with owner, editor, and viewer roles.

A user supplies a set of prompts that mirror real customer questions. AEO Copilot sends each prompt to every configured engine, reads the responses, and extracts whether the tracked brand was named, where it ranked among named brands, which sources the engine cited, and what sentiment the answer conveyed. Recommendations come back ranked by impact and tied to the specific prompt that triggered each finding. That data can be consumed as a PDF or shareable-link report, pulled programmatically through the REST API, or streamed into an AI agent over MCP using the command claude mcp add aeo-copilot -e AEO_COPILOT_API_KEY=aeo_your_key -- npx aeo-copilot-mcp.

Why it matters

The platform also exposes three no-account tools: a Free AEO Checker that runs a live ChatGPT prompt against a brand and returns six technical checks and a ranked fix list; a Free AEO Website Audit that scores whether a model can parse a URL and quote it accurately across heading structure, schema markup, quotable passages, FAQ blocks, and indexability; and a Brand vs Brand comparison that puts two brands side by side in ChatGPT answers, expandable to five competitors after sign-in. Data can be imported in bulk from a spreadsheet and exported as xlsx, CSV, or JSON. The author states that API keys are never destroyed, so they resume authenticating if a user moves from the free tier back to a paid plan.

Answer engine optimization is the practice of getting a brand named when AI tools respond to category questions that precede a purchase. Most existing tools that track brand mentions across AI engines are priced for enterprise procurement cycles, which effectively locks out freelancers and small agencies. AEO Copilot's creator makes this explicit, noting that the budget normally spent on lead generation and advertising was redirected into the free tier instead. The MCP integration is a differentiating design choice: rather than building a proprietary AEO agent, the product acts purely as a structured data source, letting users pipe findings into whatever coding or content agent they already use, including Claude Code, Codex, Cursor, and OpenCode.

Who is affected

The practical consequence for content and SEO work is that a closed loop becomes possible: AEO Copilot measures AI visibility, the connected agent reads the ranked recommendations and edits the relevant pages, and the next scheduled run measures whether the changes produced a different answer. The creator describes this explicitly as the intended workflow. Because the MCP server is a standard protocol implementation rather than a proprietary plugin, no additional marketplace approval is needed and any MCP-compatible agent can connect to it immediately.

Three audiences are called out directly. AI-native startups running their own agents are the first: they gain time-series proof that AI engines name their product, surfaced through whatever agent they already operate rather than a separate dashboard habit. SEO and marketing freelancers are the second: the white-label report branding on the Freelancer and Agency plans lets them hand a client a formatted report without exposing the underlying toolchain, and the per-brand-unlimited pricing structure on the Agency plan means adding a new client does not increase the per-seat cost. Webflow and website agencies are the third group, given direct access to track every client brand in a single account.

On the builder side, developers integrating the MCP server should verify that their agent runtime supports the npx-based MCP server pattern and that the AEO_COPILOT_API_KEY environment variable is accessible in the agent's execution context. Teams evaluating the Google Search Console integration should confirm which plan tier activates it, as the feature list notes its availability without specifying a plan threshold. The data retention and export terms state that nothing is deleted when a subscription lapses, but API keys do stop authenticating on the free plan, meaning agent-driven workflows will stall silently unless the code handles a 401 response explicitly.

What to watch next

The most concrete thing builders should test first is whether the 50-prompt monthly limit on both the free and the $9 Startup plan is sufficient to cover a realistic set of category questions. Fifty prompts spread across four engines on a paid plan means 50 prompt executions total, not 50 per engine, so the effective coverage depends on how many distinct queries the brand needs to track. The twice-monthly and weekly scheduling options on paid plans run automatically on fixed calendar dates — the first and fifteenth, or every Sunday — meaning teams that want to react to a product announcement mid-cycle will need to trigger a manual run rather than wait for the next scheduled window.

The author has published 21 SEO and AEO skills for Claude Code under the sofianbettayeb/seo-aeo-copilot marketplace, installable with the command /plugin marketplace add sofianbettayeb/seo-aeo-copilot. Whether those skills cover the full prompt-to-fix cycle or require additional configuration is worth testing before committing the workflow to a client engagement. The platform's llms-full.txt file at aeo-copilot.com/llms-full.txt is the designated onboarding document for agents, so reviewing it before connecting an agent will clarify exactly which MCP endpoints are exposed and what each one returns.