Six months ago a client forwarded me a screenshot of ChatGPT recommending three competitors in their space. Their brand wasn't on the list — despite having better reviews, more customers, and having been in the market longer than two of the three companies mentioned.

That was the moment I understood something had quietly changed about how potential customers were discovering businesses — and neither I nor my client had been paying attention.

The 10-Minute Test

Before doing anything else, run this simple diagnostic. Open ChatGPT, Perplexity, and Google AI Overviews and ask the questions your target customers most commonly ask when they're in research mode:

Log the results: Does your brand appear? Is the description accurate? Who appears instead?

Our results: The brand appeared in about 2 out of 10 queries. The main competitor appeared in 9 out of 10. When the brand did appear, the AI got the core use case wrong.

Why This Was Happening

After digging into this — and working through GEO: The New Enterprise Gateway in the AI Era by Huang Shuang, Li Jing, Wu Hao, Luo Wei, and Yuntao Zou (published June 2026, City University of Macau) — I understood the root problem.

AI systems don't read your homepage the way Google crawls it. They construct answers from whatever sources give them something structured and verifiable to work with. The book frames this as the core challenge of Generative Engine Optimization: it's not a ranking problem, it's an interpretation problem.

Our client's website was full of phrases like "industry-leading solution" and "customer-centric approach." An AI literally cannot do anything with those phrases — they contain no extractable, comparable information. The competitor's site had specific integration counts, named customer logos with industries, response time SLAs, pricing tiers by plan, and a comparison table.

What We Changed

Product and service pages

Added specific specs, use case descriptions, a clear statement of who the product is not for, pricing range, and a FAQ section written in the exact language customers use when searching.

Comparison page

Created a dedicated "[Brand] vs [Competitor A] vs [Competitor B]" page. Fact-based, not promotional — a straightforward table covering the dimensions customers actually care about.

Case studies

Rewrote all case studies to include actual numbers: customer size, specific metric improvements, exact timelines. "We helped a mid-size e-commerce company reduce support tickets by 34% in 90 days" instead of "we helped a company improve their operations."

Third-party profiles

Ensured G2 and Capterra profiles were accurate, complete, and up to date. These platforms get cited heavily by AI systems and are not optional for B2B brands.

Results at 60 Days

Mention rate across the three platforms went from roughly 20% to roughly 65% of relevant queries. Description accuracy improved significantly — the AI now mostly gets the use case right.

More tangibly: the sales team started hearing customers say "I looked you up on ChatGPT first" — which hadn't happened before. Two deals in the pipeline came in with customers who had pre-formed accurate impressions of the product before the first call.

The Mindset Shift That Helped

The framing from GEO: The New Enterprise Gateway in the AI Era was useful here. The authors distinguish between three stages of AI visibility: being seen (the AI mentions you), being understood (the AI describes you accurately), and being chosen (the AI proactively recommends you for relevant queries).

We had been focused on the wrong question — "how do we show up more?" — when the real question was "does the AI have enough accurate, structured information about us to describe us correctly when asked?" That reframe changed everything we prioritized.

If you haven't run the 10-minute test above, do it before your next client call. For the full framework, GEO: The New Enterprise Gateway in the AI Era by Huang Shuang et al. is the most rigorous treatment available — on Amazon here.