By Alexander Vinokurov
This is Part 4 of our AI search experiment. Part 1 set the baseline and the checklist, Part 2 measured the first 31 days, and Part 3 documented the layer that serves every page as Markdown to AI agents. This part reports what three months of Google Search Console data and four weeks of Google Analytics 4 data, checked on 20 September 2026, show about visibility. The commercial side — clicks, leads, page-level data and our next steps — is now reserved for clients, and we explain why below.
The numbers we publish
| Metric | Value | Period | Source |
|---|---|---|---|
| Google Search impressions | 6,316 | 19 Jun–18 Sep 2026 | Search Console, Web |
| Average impressions per month | ≈2,100 (vs 479 in the first 31 days) | 19 Jun–18 Sep 2026 | Search Console, Web |
| Impressions in Google's AI features | 259 | 19 Jun–18 Sep 2026 | Search Console, Generative AI features (Beta) |
| AI-assistant sessions vs organic search | ≈1.9× | 23 Aug–19 Sep 2026 | GA4, AI Assistant channel |
| Engaged rate of AI-assistant sessions | 63% (vs 19% for Direct) | 23 Aug–19 Sep 2026 | GA4 |
The first-month figure comes from Part 2 and overlaps the three-month window by about three weeks, so it is not a clean before-and-after — but the direction is not in doubt. The AI-feature impressions are impressions inside Google's AI answers, reported separately; we do not add them to the Web total, because the same searcher can be counted in both.
🔒 Locked: the numbers we keep for clients
| Data | Access |
|---|---|
| Organic clicks and click-through rate | 🔒 Clients only |
| Leads and inquiries by channel | 🔒 Clients only |
| Page-by-page impressions, clicks and positions | 🔒 Clients only |
| The buyer queries that reached the top 10 | 🔒 Clients only |
| What did not work, and why | 🔒 Clients only |
| Our plan for the next quarter | 🔒 Clients only |
From this part on, the commercial data stays private — not out of modesty. Page-level data is a map: it shows exactly which buyer questions, markets and page types are worth targeting, and publishing it hands that map to every agency competing for the same searches. We walk through the full dataset, and how the same approach would apply to your market, on a call or as part of an AI-SEO & AEO audit.
What we can share: four patterns
- Market-specific buyer pages beat generic ones. Pages that answer one buyer question in one market earned most of the impressions — far more than broad service pages.
- Comparison content gets picked up by Google's AI features. Pages that honestly compare providers were among the most shown inside Google's AI answers.
- AI assistants send people to proof. Most AI-assistant visits landed on case studies, not on blog posts about case studies — so case studies have to be written for machines as well as for people.
- Visitors from AI assistants arrive qualified. They engaged at more than three times the rate of direct traffic, because the assistant has already explained who we are and what we do.
What did not matter: special files for Google. Google's own guidance on AI features says sites do not need new machine-readable files, AI text files, markup or Markdown to appear in its generative AI results, and that no special schema is required. We keep llms.txt and Markdown because other agents use them — the gains above came from clearer pages, not from technical tricks.
What three months do not prove
Three months shows direction, not causation. Several changes shipped together, GA4 and Search Console measure different windows, and the AI Assistant channel includes some of our own checks. What the data does show is that clear, market-specific buyer pages get seen by Google and its AI features, and that AI assistants send engaged visitors straight to case studies.
Get the same system for your company
The playbook behind these results is what we sell as AI-SEO & AEO: an entity and topic map, answer-ready buyer pages for the markets you sell into, case studies structured so AI assistants can cite them, and an internal link system that routes that attention to contact. A one-off audit is a fixed $1,500; ongoing support — shipping the pages, links and structured content — starts from $1,500 a month.
For the baseline and checklist, start with Part 1; for the first-month numbers, read Part 2.
Frequently asked questions.
Want the full numbers and the playbook?
We share the locked data from this experiment with clients — and apply the same system to your site: entity clarity, answer-ready buyer pages, AI-citable case studies and internal links that lead to contact. Audit $1,500, ongoing support from $1,500/month.
