Where is the Google Search Console for LLMs?

§ The problem: numbers nobody can verify

Here's a scene playing out in marketing meetings everywhere this year. A business pays for an AI-visibility report. The score says 62. A month later — same website, same content, nothing changed — the score says 41. The vendor shrugs: "AI answers fluctuate."

So which number was true?

Both. Neither. Nobody can say — and that's the actual problem.

If this feels familiar, it should. SEO rankings have always bounced around — up one month, down the next. But SEO had something AI visibility doesn't: a referee. When the numbers were disputed, you opened Google Search Console and looked at Google's own data — impressions, clicks, queries, straight from the source. Third-party rank trackers could disagree all they wanted; the console settled it.

AI search has the volatility — with no referee.

§ The evidence: the industry is losing patience

This isn't a quiet complaint anymore. In technical SEO communities this month, practitioners running real client accounts are saying it out loud: they no longer trust AI-visibility dashboards, because the numbers are estimates built by sampling prompts — and the estimates won't sit still. One thread drew dozens of practitioners comparing notes on the same frustration: clients ask "is ChatGPT actually reading our site?" and the tools can't answer it.

The distrust has an economic backdrop: money is pouring into AI-visibility tooling faster than measurement standards can form. G2 launched its first AEO software category less than a year ago and watched it grow over 2,000% — dozens of dashboards, all promising to tell you what AI says about your brand, none able to show their receipts. Meanwhile the deeper studies keep confirming why the numbers wobble: the same question, asked of the same engine, returns different answers on different runs — and increasingly, different answers for different users, as engines personalize with memory.

None of this makes the tools useless. It makes them polls. And the industry has been selling polls priced and presented as measurements.

§ Why is this actually an issue? Because the engines publish (almost) nothing

Blaming the tools misses the real story. Third parties are guessing because the engines have left everyone no choice. Look at what the platforms actually provide to the businesses they talk about every day:

That's the scoreboard: one partial console, one blended report, two silences.

§ This is a policy-level gap — and SEO already lived through it

Call it what it is: not fraud, not incompetence — missing infrastructure. The engines have no obligation to report, no standard to follow, and no pressure yet strong enough to make them. Early SEO looked exactly like this: years of third-party guesswork, disputed numbers, and snake oil — until Google shipped the console's ancestor in the mid-2000s and the discipline professionalized almost overnight. First-party data didn't just settle arguments; it separated real practitioners from pretenders.

The same correction is coming to AI search, and you can already see the forcing function: advertising. Now that ChatGPT sells ads, advertisers will demand attribution — they always do — and measurement infrastructure built for advertisers has a long history of eventually surfacing for organic visibility too. Someone will define the standard for AI-citation reporting; the only question is when, and whose vocabulary it uses.

Until then, every business is operating in the gap — which is exactly why how you measure right now matters more than which tool you buy.

§ What can you actually do today? Five honest options

1. Use visibility tools as trendlines, never as truth. Fix your question set, fix your schedule, and watch the direction over months — ignore the absolute score. A poll can't tell you your exact standing; it can tell you if you're rising.
2. Use the one first-party report that exists. Bing Webmaster Tools' AI Performance report is free and takes ten minutes to set up. It covers one ecosystem — but it's the only place any engine currently shows you receipts.
3. Ask your web host one question: "which AI services are visiting our site, and which pages?" Your website keeps its own records of every visit — including AI engines reading your pages. Think of them as read receipts. Most businesses have never looked at theirs.
4. Deputize your front desk. Add one question to every intake call and form: "How did you hear about us?" — and count the answers that say ChatGPT, Gemini, or "an AI." It's the crudest metric on this list and, for connecting AI visibility to revenue, still the most honest one.
5. Triangulate. No single number can be trusted alone right now — but a trendline, a Bing report, your site's read receipts, and your intake tally all pointing the same direction? That's as close to truth as 2026 offers. Three imperfect signals agreeing beat one confident score.

§ Frequently asked questions

FirePencil.AI is an autonomous AEO agent. This article is general information about measuring AI visibility, not professional advice. Third-party names (ChatGPT, Gemini, Perplexity, Claude, Google AI Overviews, Bing Copilot, Google Search Console, Bing Webmaster Tools) are trademarks of their respective owners; use is descriptive and does not imply affiliation or endorsement.