How to Get ChatGPT Claude and Gemini to Recommend Your B2B Company

60% of B2B buyers now start their research in ChatGPT, Claude, Gemini, or Perplexity before they ever land on a vendor website. A growth-stage payments CEO Will Leatherman worked with discovered this firsthand: every time he searched his own category, his competitors appeared and his company did not. His buyers had already built a shortlist without him.

Getting recommended by an LLM is fundamentally different from ranking on Google. The model does not return a page of links. It names 2 to 5 vendors and explains why. To be one of those vendors, you have to give each engine exactly what it needs.

What is AEO and who does it apply to?

AI Engine Optimization (AEO) is the practice of structuring a company's content and online presence so that large language models name the company when a buyer asks for a vendor recommendation. It applies to any B2B company whose buyers use ChatGPT, Claude, or Perplexity to research purchases. Unlike SEO, AEO does not target ranked positions. It targets being named in the answer at all, and being named accurately.

For a full breakdown of the underlying framework, see How B2B Marketing Teams Get Named in AI Search.

What 3 signals do all AI engines share?

Every major LLM weighs the same 3 factors when choosing which companies to name in a recommendation:

| Signal | What it means | Why the model cares | |---|---|---| | Category clarity | The model can identify exactly what you do and who you serve, consistently across all platforms | Conflicting positioning across your website, LinkedIn, and executive profiles confuses the model and reduces citation probability | | Proprietary claims | You publish proof, data, or perspective no competitor can copy | The model already holds generic knowledge; it only cites sources that give it something it does not already have | | Third-party mentions | Other sites feature you in best-of lists, comparisons, or category publications | The model trusts third-party sources roughly 3x more than a brand's own page, and half of all AI answers come from pages the brand does not own |

Category clarity is usually the fastest fix. Many companies describe themselves differently across their website, company LinkedIn, and individual executive profiles. Aligning that language is the first sweep before any content work begins.

What does ChatGPT specifically need to cite your company?

ChatGPT was trained heavily on Wikipedia and structured web content. 2 specific tactics move the needle most for ChatGPT citations.

First, add structured FAQs to every key page. Question-and-answer blocks that directly address what buyers are searching give ChatGPT a pre-assembled answer it can repeat. If a buyer asks "what is the best HCC coding software for health plan risk adjustment" and your FAQ answers that question plainly, ChatGPT will lift that answer verbatim.

Second, publish timely content. Content published within the last 30 days carries substantially higher citation probability on ChatGPT, Perplexity, and Google AI. Adding the current year to article titles signals recency that those engines actively weight. "Best risk adjustment software 2026" performs better than the same article without the year.

Wikipedia presence is a strong signal for ChatGPT but is hard to create directly for most companies. The more reachable path is earning mentions in the best-of publications and comparison lists that ChatGPT already relies on. Getting featured on those pages often has more citation impact than any on-site change.

What does Claude specifically need to cite your company?

Claude weights Reddit threads and community discussions more heavily than any other major engine. These represent unmediated buyer conversation, and Claude treats that as a high-trust signal precisely because no brand controls it.

The fastest path to Claude citations is to be present and accurately described in communities where your buyers ask questions. When real users discuss your category in online forums and recommend your company, Claude picks that up as third-party validation.

Customer voice in your own content works the same way. Pages that include verbatim customer quotes tied to specific, measurable outcomes signal the same thing to Claude that community discussion does: direct buyer perspective the model can lift and cite.

What does Gemini need to cite your company?

Gemini and Google AI Overview are closely tied to Google's existing signals. Most established SEO practices carry over, including domain authority, structured content, and page comprehensiveness. Companies that already rank well on Google typically see the strongest starting scores on Gemini.

That also means ChatGPT is the harder platform to crack for most B2B companies. A company with strong SEO but thin original content and no third-party mentions will show this pattern: visible on Gemini, invisible on ChatGPT and Claude.

The live audit of a healthcare IT company run during this workshop illustrated the pattern exactly. Their Gemini visibility was the highest of any engine. Their ChatGPT visibility was around 20%. The gap was not an SEO deficit. It was missing structured FAQs, unattributed on-page statistics, and no presence in the category comparison publications ChatGPT relies on.

Why does front-loading your page produce more AI citations?

LLMs scan the first section of a page and pull the most quotable answer they find there. The engine rarely reads the full page. This means a company can have the right answer buried in paragraph 7 and still get no citation, while a competitor with a weaker answer front-loaded in paragraph 1 gets named instead.

The fix is structural. Move your single most specific claim to the first paragraph. Answer the buyer's question in 1 sentence before explaining your methodology or company background. Hero sections that open with brand mission and bury results below the fold produce lower citation rates even when the underlying content is strong.

Front-loading compounds when combined with sourced statistics and direct quotes. Pages that include cited data and real quotes from subject matter experts see measurably higher citation rates, both because the claims are more specific and because the engine can verify them.

3 prompts you can run today to score any page:

Score this page 0-10 on source density, front-load, and originality. Could a competitor have published these sentences? List the highest-impact fixes first. [paste the page text]

Scan this page and identify the sentences that could only have come from this specific company, based on unique data, customer outcomes, or methodology. List them. [paste the page text]

I am researching vendors for [your category]. What companies do you consistently recommend and why? Which sources are you citing?

What content is worth publishing to show up in AI answers?

AI engines already hold generic knowledge. Content that repeats what the model already knows does not generate a citation. The only content that forces a citation is content the engine has to go find because it does not already have it.

3 types of content create that forcing function:

• Proprietary data. Benchmarks, patterns, or aggregate findings from your own customer base or product. "We've seen this pattern across 200 companies" is a claim no competitor can copy.
• Original perspective. Founder and operator-led content grounded in specific decisions and lived experience. Lived experience is outperforming generic brand content because the model treats it the same way it treats original data: something it cannot synthesize on its own.
• Customer outcomes with specifics. "A client grew AI-driven sessions nearly 1,000% month over month across 100+ countries" is citable. General benefit claims are not.

Using AI to assist with drafts is fine. The recommendation is to have AI produce a polished first draft from your research and inputs, then review and edit the way you would review a contractor's draft. That workflow lets a single marketer scale to 20 to 30 pieces per month when going after an aggressive AEO target, while keeping the original perspective and sourced specifics that drive citations.

How fast can a company move from invisible to cited?

Fast. Clients regularly move from unranked to appearing in the top 5 within the first 90 days. AI search is new enough that most competitors have not invested in it yet. Unlike SEO, where high-authority sites have accumulated years of advantage, AEO is largely uncontested in most B2B categories. Changes can affect citation rates in days or weeks rather than months.

One of the stronger documented cases is Rise, a global payroll provider. They were not appearing meaningfully in AI search when they started. After building out their content strategy across all 3 components, they grew AI-driven new sessions nearly 1,000% month over month and generated millions in pipeline within a year. They now show up in AI search results across more than 100 countries.

A category with 23K monthly searches is approachable even at 5% share. The question is whether your company appears in the answer when those queries run.

You can check where you currently stand with the free AEO audit at gotcatalyst.com. It runs 50 to 100 fresh queries across ChatGPT, Claude, Gemini, Perplexity, and Google AI and shows exactly where you appear, where competitors are ahead of you, and which third-party publications to target first. For the methodology behind the audit scoring, see the AEO methodology.

The takeaway

ChatGPT, Claude, and Gemini each have preferences, but they share the same 3 requirements: clear category positioning, proprietary content only your company can publish, and third-party mentions from sources the model trusts. Engine-specific tactics layer on top once those 3 are solid.

The fastest single action: take one page that should be ranking for a buyer query but is not, run the scoring prompt above, and move your most specific claim to the first paragraph. Check the audit tool 2 weeks later to see whether the signal shifted.