I typed a niche consulting keyword into Google AI Mode last week.
The AI-generated answer listed five companies. Four were global names — the kind with thousands of employees and offices on every continent.
The fifth was a boutique consultancy I had just finished auditing. Fewer than 20 people. One office.
Same answer. Same sentence. Same recommendation.
Their website scored low on my GTM audit. Weak messaging. No clear CTA. Service menu that read like a Wikipedia index.
But AI did not care about any of that.
AI was picking up a different set of signals.
The signal most B2B companies are missing
Here is what is happening: 94% of B2B buyers now use AI in their research process. Google AI Mode appears on 48% of searches. For B2B tech queries, that jumps to 82%.
When a buyer asks "who are the best consulting firms for X" — they are not scrolling through ten blue links. They are reading an AI-generated answer that names 3-5 companies.
If your company is not in that answer, the buyer never reaches your website.
This is not an SEO problem. SEO optimises for rankings. AI does not rank. AI recommends.
The signals that make AI recommend you are structurally different from the signals that make Google rank you. I have been studying what those signals are — and running checks on every company I audit.
The pattern is clear. AI cites companies that do four things.
The 4 dimensions of the AI Visibility Signal
→ AI Citation Presence — Does AI mention your company by name when a buyer searches your category keyword? Not in a footnote. Not in a linked source. In the actual answer. The boutique firm appeared in the text. Their larger competitor with better SEO did not.
→ Entity Recognition — Does AI know what your company is? When you ask "What does [Company] do?" — does it return accurate information, or does it confuse you with another entity? Companies with named methodologies and consistent descriptions across multiple sources score highest here.
→ Content Structure — Is your content built for AI to extract? Headers, bullet points, question-answer format, named frameworks, original data. AI pulls from content it can parse cleanly. Walls of text get skipped. Blog posts from 2023 get ignored.
→ Citation Breadth — How many AI platforms can find you? Google AI Mode, Perplexity, ChatGPT — each one pulls from different sources with different weights. A company that appears on one platform but not the others has a narrow signal. The strongest companies appear across all three.
These are not vanity metrics. They are buying signals in reverse — signals that tell you whether buyers can find you through the channel that is replacing search.
Why your competitor might already have this signal
The boutique firm I discovered did not optimise for AI. They did not even know they were being cited.
But they had done three things — by accident — that triggered the signal:
Their content answered specific questions that buyers were asking (not "about us" copy)
They had a named methodology that AI could reference as a distinct concept
Other sites mentioned them by name in industry context — directories, partner pages, press coverage
That was enough. Three structural signals. No AI strategy. No prompt engineering. No expensive tools.
Their competitor — a larger firm with a better website, stronger SEO, and more backlinks — was invisible. AI did not cite them once.
Because AI does not care how good your website is. It cares whether you are a referenceable, citable entity in your category.
What to do about it
I built a free scorecard that lets you check your own AI Visibility Signal in 5 minutes. Four dimensions. Self-assessment. No tools required.
Download it here: AI Visibility Signal Scorecard
If you want the full analysis — I will run your company through our scanner, score all 4 dimensions with real data, and deliver a branded report with a prioritised fix list. 48 hours. £297.
Reply to this email with your company URL. I will send the details.
One more thing
I also wrote a cheat sheet: the 7 structural changes that make AI cite your company. It is the implementation guide that pairs with the scorecard — what to fix and why it works.
Subscribers get it free. It is attached to this email as a PDF.
If you know a founder who is still optimising for Google rankings while AI is recommending their competitor — forward this issue. They need to see the signal.
The GTM Signal is a weekly diagnostic from GTM Signal Studio. Each issue identifies one signal that separates companies getting found from companies getting forgotten.