Client case study · Nutrition and diet · August to November 2026

A Nutrition Website Roughly Quadrupled Its Daily Google AI Impressions in Three Months

A nutrition and diet website hired me for the full engagement: audit, action plan and implementation. Over three months, its impressions in Google’s AI features went from about 120 a day to 492 on November 15. The client asked that neither the business nor the website be named, so this page leaves out anything that would identify them.

~4×
daily impressions in Google's AI features, mid-August to mid-November 2026
492
impressions on Nov 15, 2026, up from about 120 a day in mid-August
23K
impressions in Google's AI features, Aug 16 to Nov 15, 2026

Source: the client’s Google Search Console, Generative AI features report (Beta), web search, 3-month view, read November 17, 2026.

What the Number Measures

Google Search Console’s Generative AI features report counts how often a site appeared in Google’s AI surfaces, such as AI Overviews and AI Mode. For this site the count was about 120 a day in mid-August. It more than doubled by mid-September, held between about 200 and 320 a day through October, then climbed again in November to 492 on November 15.

Nutrition is a category where people ask detailed questions and AI answers are cautious about which sources they draw on. Being one of those sources is the starting point for being trusted and chosen.

Google Search Console Generative AI features report for an anonymized nutrition and diet client, web search, three months from August 16 to November 15, 2026: 23K total impressions, daily impressions rising from about 120 to about 490.
The client’s Google Search Console, Generative AI features report, web search, August 16 to November 15, 2026. Cropped so the site cannot be identified.

The Engagement: Audit, Plan, Implementation

Unlike an audit-only project, this one ran end to end.

STEP 1
Audit
Where the site stood in AI answers for the questions its readers ask, which competitors and sources were cited instead, and what on the site was holding it back.
STEP 2
Action plan
A prioritized list of fixes, ordered by expected impact and effort, so the most valuable changes shipped first.
STEP 3
Implementation
The plan was carried out, not just handed over: technical fixes, content, AEO and GEO work, delivered and checked on the live site.

What Was Worked On

Four workstreams, delivered over the same three months.

Technical
Making sure the pages that should be read are crawlable, fast to parse and clearly described, with structured data that matches what is on the page.
Content
Pages built or reworked so each one answers a specific nutrition question clearly and completely, in the order a reader needs it.
AEO
Answer-first structure: the direct answer at the top of the page, followed by the detail, so an AI system can lift a correct answer and credit the source.
GEO
Strengthening the signals generative engines use to decide which sources to trust and blend into longer answers, on the site and beyond it.

What This Does Not Show

  • Clicks or revenue. The report counts impressions only, with no clicks, click-through rate or queries, so this page makes no traffic or revenue claim.
  • Which surface. AI Overviews and AI Mode are reported together.
  • Proof of cause. Google kept expanding its AI features during these months, which lifts many sites at once. There was no control site, so read this as a measured result, not proof that each change caused it.

Find Out Where Your Site Stands

The free check shows how ChatGPT, Gemini and Google answer the questions your customers ask. The full audit adds the reasons and a prioritized plan. Scoped to your business, confirmed on a free call.

How This Was Measured

  • Client: a nutrition and diet website. Business and site names withheld at the client’s request.
  • Source: Google Search Console, Performance, Generative AI features (Beta), search type Web (text), 3-month view covering August 16 to November 15, 2026, read November 17, 2026.
  • Figures:23K total impressions and 492 on November 15 are the console’s own values. The mid-August level (about 120 a day) and the October range are read from the chart and are approximate.
  • Impressions only. Figures as measured; they show association, not proof that any single change caused them, and are no guarantee of similar results.

More results: all case studies.