Why ChatGPT Names Your Competitor Instead of You and the 2 Free Checks to Fix It

51% of B2B software buyers now start their search in ChatGPT, Claude, or Perplexity. That number was 29% just 11 months ago. The speed of that shift means most companies are still optimizing for a buyer behavior that no longer describes the majority of their pipeline.

Catalyst builds AI visibility systems for B2B companies. In this workshop, Will Leatherman walked through why AI answers are wrong, which popular fixes waste your time, and the two checks any team can run for free tonight.

What is AI answer engine optimization and why does it matter now?

Answer engine optimization (AEO) is the practice of influencing what AI chatbots say about your company when a buyer asks a question in your category. It is different from SEO. Only 8% of ChatGPT citations rank in Google's top 10 for the same prompt. Optimizing for one does not reliably move the other.

The stakes are real. A solar company in Minnesota lost a $150,000 contract after Google AI told a buyer the company was being sued by the state attorney general. They never were. Four real citations sat underneath that answer. None of them said it. There is no correction form at OpenAI, Google, Anthropic, or Perplexity. You cannot file a ticket. The only path to a different answer is changing what the sources say.

Catalyst went from zero AI citations to category leader recommendation within a few months using the same approach described in this workshop.

Why is ChatGPT confident but completely wrong about your company?

Two mechanics explain most of the inaccuracy.

First, LLMs rewrite your prompt before retrieving anything. When a buyer types "best CRM for B2B SaaS" into ChatGPT, the model fans out into a cluster of related queries before it fetches a single source. It might run "2026 best CRMs", "HubSpot vs. Pipedrive", and "automated CRM sync for SaaS" in parallel, then aggregate the results into one answer. Testing your own prompt once tells you almost nothing about what a buyer with different conversation memory will see, because the model folds prior context into how it rewrites the query.

Second, the model builds an entity profile from third-party sources, not your own pages. Across 40 B2B SaaS categories and 233 recommendations, ChatGPT cited the tool's own website only 11.6% of the time and cited G2 or Capterra zero times. Most citations come from roundups, best-of lists, industry blogs, and publication features. The product parameters those third parties use to describe you explain 82.4% of the variation in whether you get recommended. Your brand identity explains 1.2%.

Mud Ventures rebranded and 3 months later ChatGPT was still using their old company name and a retired domain, even with a working 301 redirect in place. Seven weeks of Wikipedia edits, schema markup, press releases, and Reddit posts moved nothing. Entity retraining only happens at the next model update. Getting accurate information into third-party sources early is the only move that isn't a waiting game.

Which popular fixes are wasting your time right now?

Before spending a single hour on tactics, know what the data says does nothing.

What is actually blocking AI from reading your site?

Most companies have an infrastructure problem before they have a content problem. Fix this first.

In July 2025, Cloudflare made AI bot blocking the default setting for new domains. That change affected roughly 20% of the web. Your site may be serving a 200 OK to the ChatGPT bot in your robots.txt while a Cloudflare firewall rule returns a 403 on every actual request. Both things can be true at the same time. A robots.txt check is not enough.

A scan of 500 million ChatGPT bot fetches found that the bot executes zero JavaScript. If your site renders on the client side, the bot gets an empty shell. Google's crawler renders JavaScript fine, so your SEO tool shows a healthy page while the AI crawler sees nothing.

Check your Cloudflare and hosting settings. Confirm that the ChatGPT user-agent passes through your WAF rules, not just your robots.txt. If your site is client-side rendered, work with your engineering team to ensure critical pages return server-rendered HTML.

What does ChatGPT actually pull citations from?

Off-site authority drives AI citations more than any on-site change you can make. The sources LLMs trust most are the same ones buyers consult when they want a second opinion: industry roundups, best-of lists, category comparison articles, LinkedIn, and niche publications.

The Catalyst AEO audit maps exactly which of those sources ChatGPT, Gemini, and Perplexity are pulling for queries in your category, and which ones currently name a competitor instead of you. For Attio, a modern CRM, the audit showed 10 best-of roundups actively cited in AI answers. Attio appeared on 1 of the 10. The other 9 are outreach targets with a clear business case: get featured, get cited.

LinkedIn shows a particularly strong correlation with AEO citations. Consistent, on-topic content on LinkedIn feeds the LLM's entity model for your company and the people who run it.

How do you move your AI visibility in a reasonable timeframe?

Three concrete actions from the workshop:

1. Fix the infrastructure first. Confirm AI bots can actually fetch your pages. Check Cloudflare rules, WAF settings, and whether your site serves real HTML to bots without JavaScript. This takes a few hours with an engineer and can unlock visibility that no content work can compensate for.

2. Build a file of truth. Consistent messaging across your website, newsletter, LinkedIn, and press is the foundation AI uses to build its entity model of your company. If your site calls you a founder, your LinkedIn calls you a CEO, and a podcast says you run a startup, the model struggles to resolve who you are. Write a single-source definition of what you do, who you serve, and how you describe your product, then publish it consistently everywhere. Every time you update your positioning, update the file.

3. Earn placements on the roundups that matter. Identify the best-of lists and comparison articles that show up in your category's AI answers. Find the authors. Pitch for inclusion with specific evidence that your tool belongs on the list. This is the highest-leverage move in the off-site category and it compounds: getting featured on one list increases the probability of being pulled into AI answers, which can prompt editors of other lists to add you.

The AEO audit at Catalyst runs this analysis automatically. It returns the exact list of publications currently shaping your AI presence, sorted by citation frequency, with author information so you can begin outreach directly. It is free to run.

For more on which page types drive AI citations, see The 4 Types of Pages That Drive AI Citations. For a step-by-step method for tracking how your brand ranks across chatbots, see How to Track Your Brand Ranking Across AI Search Engines. Catalyst's AEO visibility service builds and runs these systems end to end for B2B companies.

The takeaway

AI chatbots are now the first stop for 51% of B2B software buyers. The model's answer comes almost entirely from third-party sources, and you have no form to submit corrections. Tactics that dominate the conversation, such as LLMs.txt and structured data, have no measurable effect. What moves the needle is clearing the infrastructure blockers that prevent AI from reading your site, publishing consistent messaging everywhere your entity appears, and earning placements on the roundup articles and comparison pages LLMs cite most.

Run the free AEO audit tonight. It takes about 10 minutes and shows exactly where your AI visibility stands and which off-site sources to go after first.