AEO & GEO skill 05 of 13

Google AI Overviews and AI Mode SKILL.md

The Google AI Overviews and AI Mode skill makes Claude, Codex or another AI agent work out where a site appears in Google’s AI features, why it doesn’t appear where it should, which controls exist and how to measure it. It uses Google’s documented positions, the Search Console Generative AI performance report and independent click studies, and focuses on the clicks that remain.

ai-overviews-ai-mode/SKILL.md, version 1.0.0, 1,470 words.

AI Overviews put a generated summary above Google's results for many searches; AI Mode is a conversational search experience where most of the page is the answer. Both use retrieval-augmented generation and query fan-out over Google's Search index, and both are rooted in Google's core ranking and quality systems. [Confirmed, Google Search Central] Google reported more than 2.5 billion monthly users for AI Overviews and more than 1 billion for AI Mode in June 2026 (as reported by industry press).

This skill works out where a site appears in Google's AI features, why it doesn't appear where it should, what controls exist and how to measure it with Google's own data.

What Google has documented

TopicGoogle's positionLabel
EligibilityThe page must be indexed and eligible to show with a snippet, and the site must be included in generative AI features in Search ConsoleConfirmed
Ranking basisAI features are rooted in core Search ranking and quality systems; SEO best practices still applyConfirmed
How answers are builtRAG (grounding) and query fan-outConfirmed
Special files or markupllms.txt and special markup aren't used; structured data isn't required for AI featuresConfirmed
ContentUnique, non-commodity, people-first content matters mostConfirmed
MeasurementGenerative AI performance report in Search Console shows impressions in AI Overviews and AI Mode; launched 3 June 2026, available to all sites by 31 August 2026Confirmed
Opting outA Search Console control (tested from June 2026) can exclude a site's content from AI features; excluded sites get no traffic or impressions from those features, with no effect on other rankingsConfirmed (rollout status changes)
Google-ExtendedControls some Gemini uses; doesn't affect Google Search or AI OverviewsConfirmed

What the click data shows

StudyFinding
Pew Research Center, July 2025 (900 US adults, 68,879 searches)People clicked a traditional result on 8% of visits with an AI summary versus 15% without; they clicked a link inside the summary on 1%
Ahrefs, April 2025 (300,000 keywords)Position-one click-through rate about 34.5% lower when an AI Overview appeared
Seer Interactive, September 2025Organic CTR fell sharply on AI Overview queries, but brands cited in the overview had higher CTR than brands not cited (correlation, not proven cause)

The takeaway: informational clicks are harder to earn, and being cited is better than not being cited. Measure your own data rather than assuming industry averages.

Operating principles

  1. Win the underlying searches. Citations come from pages that rank for the main query or its fan-out sub-queries.
  2. Measure with Google's report first. Impressions in AI features by page show where you already appear.
  3. Use controls deliberately. nosnippet, data-nosnippet and max-snippet limit what can be shown, including in AI features; opting out removes AI feature visibility entirely. Know the trade-off.
  4. Optimise for the click that remains. Visitors from AI features often arrive further along; give them depth, tools and next steps.
  5. Don't chase hacks. Google names chunking, AI-only rewriting, llms.txt and inauthentic mentions as things you can ignore for Search.

When to use, and when not to

Use it when AI Overviews appear for your important queries, when clicks fell while rankings held, when deciding whether to use snippet controls or the opt-out, and when setting up AI feature reporting.

Don't use it for other engines (use chatgpt-search-visibility or answer-engine-platforms) or for general content quality work (use answer-ready-content).

Inputs to ask for, in one message

  1. Search Console access, including the Generative AI performance report.
  2. The important queries and pages, by market.
  3. Click and impression trends for those pages over 16 months.
  4. Any snippet controls in use (nosnippet, max-snippet, data-nosnippet).
  5. Business model: does the site need clicks (publisher, lead gen) or brand exposure, or both?

Raw SKILL.md file

.md

This is the exact file agents read, including the credit header and licence notice. Save it as ai-overviews-ai-mode/SKILL.md.

---
name: ai-overviews-ai-mode
description: "Diagnose and improve visibility in Google AI Overviews and AI Mode: eligibility, retrieval and query fan-out, what gets cited and why, snippet controls and the Search Console AI opt-out, the Generative AI performance report, click impact from Pew, Ahrefs and Seer studies, images, video, local and shopping, and the post-click experience for visitors who still arrive. Use whenever someone mentions AI Overviews, AI Mode, SGE, Google AI answers, clicks falling while rankings hold, being cited in AI Overviews, the Search Console Generative AI report, nosnippet, or opting out of Google AI features."
license: CC-BY-4.0
metadata:
  author: Afeez Adebayo
  author_url: https://afeezadebayo.com
  author_linkedin: https://www.linkedin.com/in/afeez-adebayo/
  author_x: https://x.com/za_icon
  source: https://afeezadebayo.com/skills-md/marketing/aeo-geo/ai-overviews-ai-mode/
  category: marketing
  subcategory: aeo-geo
  version: "1.0.0"
  created: "2026-10-10"
  updated: "2026-10-10"
---

# Google AI Overviews and AI Mode

AI Overviews put a generated summary above Google's results for many searches; AI Mode is a conversational search experience where most of the page is the answer. Both use retrieval-augmented generation and query fan-out over Google's Search index, and both are rooted in Google's core ranking and quality systems. [Confirmed, Google Search Central] Google reported more than 2.5 billion monthly users for AI Overviews and more than 1 billion for AI Mode in June 2026 (as reported by industry press).

This skill works out where a site appears in Google's AI features, why it doesn't appear where it should, what controls exist and how to measure it with Google's own data.

## What Google has documented
| Topic | Google's position | Label |
|---|---|---|
| Eligibility | The page must be indexed and eligible to show with a snippet, and the site must be included in generative AI features in Search Console | Confirmed |
| Ranking basis | AI features are rooted in core Search ranking and quality systems; SEO best practices still apply | Confirmed |
| How answers are built | RAG (grounding) and query fan-out | Confirmed |
| Special files or markup | llms.txt and special markup aren't used; structured data isn't required for AI features | Confirmed |
| Content | Unique, non-commodity, people-first content matters most | Confirmed |
| Measurement | Generative AI performance report in Search Console shows impressions in AI Overviews and AI Mode; launched 3 June 2026, available to all sites by 31 August 2026 | Confirmed |
| Opting out | A Search Console control (tested from June 2026) can exclude a site's content from AI features; excluded sites get no traffic or impressions from those features, with no effect on other rankings | Confirmed (rollout status changes) |
| Google-Extended | Controls some Gemini uses; doesn't affect Google Search or AI Overviews | Confirmed |

## What the click data shows
| Study | Finding |
|---|---|
| Pew Research Center, July 2025 (900 US adults, 68,879 searches) | People clicked a traditional result on 8% of visits with an AI summary versus 15% without; they clicked a link inside the summary on 1% |
| Ahrefs, April 2025 (300,000 keywords) | Position-one click-through rate about 34.5% lower when an AI Overview appeared |
| Seer Interactive, September 2025 | Organic CTR fell sharply on AI Overview queries, but brands cited in the overview had higher CTR than brands not cited (correlation, not proven cause) |

The takeaway: informational clicks are harder to earn, and being cited is better than not being cited. Measure your own data rather than assuming industry averages.

## Operating principles
1. **Win the underlying searches.** Citations come from pages that rank for the main query or its fan-out sub-queries.
2. **Measure with Google's report first.** Impressions in AI features by page show where you already appear.
3. **Use controls deliberately.** `nosnippet`, `data-nosnippet` and `max-snippet` limit what can be shown, including in AI features; opting out removes AI feature visibility entirely. Know the trade-off.
4. **Optimise for the click that remains.** Visitors from AI features often arrive further along; give them depth, tools and next steps.
5. **Don't chase hacks.** Google names chunking, AI-only rewriting, llms.txt and inauthentic mentions as things you can ignore for Search.

## When to use, and when not to
Use it when AI Overviews appear for your important queries, when clicks fell while rankings held, when deciding whether to use snippet controls or the opt-out, and when setting up AI feature reporting.

Don't use it for other engines (use `chatgpt-search-visibility` or `answer-engine-platforms`) or for general content quality work (use `answer-ready-content`).

## Inputs to ask for, in one message
1. Search Console access, including the Generative AI performance report.
2. The important queries and pages, by market.
3. Click and impression trends for those pages over 16 months.
4. Any snippet controls in use (`nosnippet`, `max-snippet`, `data-nosnippet`).
5. Business model: does the site need clicks (publisher, lead gen) or brand exposure, or both?

## Workflow

### Step 1. Find where AI features appear
Check the target queries in the target market and record whether an AI Overview appears, which sources it cites and whether AI Mode answers cite you.

### Step 2. Read the Generative AI report
List pages with AI feature impressions. Compare with the overall Performance report to see which pages gained or lost clicks as AI impressions grew.

### Step 3. Diagnose missing citations
For important queries where you rank but aren't cited, compare your page with cited pages: direct answer, specificity, evidence, freshness, visual content. For queries where you don't rank, the issue is classic SEO first.

### Step 4. Cover the fan-out
Use `query-fan-out-mapping` to find sub-queries the overview draws on and strengthen sections that answer them.

### Step 5. Review snippet controls
Check templates for unintended `nosnippet` or very small `max-snippet` values that block use in AI features. Use `data-nosnippet` only for text that must never be quoted.

### Step 6. Decide on the opt-out
Only consider excluding the site if AI features clearly cost more than they return (for example, content licensing businesses). Model the lost impressions first.

### Step 7. Improve the post-click experience
For pages cited in AI features, make sure visitors find depth, tools, data or next steps the summary couldn't give.

## Images and video
Google notes AI features can include relevant images and video. Image and video SEO best practices therefore also support AI feature visibility. [Confirmed]

## Local and shopping
For local businesses and retailers, Google Business Profile and Merchant Center feeds help products and services appear in AI responses. [Confirmed] See `ai-commerce-product-visibility`.

## Worked example
A personal finance publisher's clicks for "how to calculate stamp duty" fell 40% year on year while average position stayed at 2. The Generative AI report showed the page with high AI impressions; the overview cited a government page and a competitor calculator. The publisher rebuilt its page around an interactive calculator with worked examples for first-time buyers and second homes, a direct answer passage and a dated rate table. The page became one of the cited sources, and clicks recovered part of the loss, with longer sessions and more calculator use than before.

## Output format
```
# AI Overviews and AI Mode review: <site>, <market>, <date>
## AI feature presence for target queries (appears, cited sources, us cited?)
## Generative AI report summary (pages, impressions, trend)
## Missing-citation diagnosis by query
## Controls review (snippet controls, opt-out decision)
## Actions with owners and re-check dates
```

## Snippet controls and their effects
| Control | Effect on AI features | Use when |
|---|---|---|
| nosnippet | Page not used in snippets or AI features | Content must not be quoted at all |
| data-nosnippet on an element | That element excluded | Protecting specific text such as prices or gated content |
| max-snippet:[n] | Limits snippet length | Balancing visibility and control |
| Search Console generative AI setting | Opts the site out of AI features | After modelling lost impressions |
Check templates for accidental nosnippet tags; they quietly remove pages from AI Overviews.

## Common mistakes
- Blaming AI Overviews for drops caused by ranking or technical issues.
- Using `nosnippet` sitewide and wondering why you're never cited.
- Opting out without modelling lost visibility.
- Thinking Google-Extended controls AI Overviews.
- Expecting AI-specific clicks in Search Console; the report shows impressions.

## Questions to ask before you recommend changes
- Does the business need clicks, exposure, or both from these queries?
- Which pages have AI feature impressions, and are clicks rising or falling there?
- Are there legal or licensing reasons to limit snippets?

## Self-check before you answer
- Did I separate Google's documented positions from industry studies?
- Did I use the Generative AI performance report?
- Did I check snippet controls and explain the opt-out trade-off?
- Are recommendations grounded in ranking for the main and fan-out queries?

## Official documentation and research
- Google: generative AI features in Search, site owner guide: https://developers.google.com/search/docs/fundamentals/ai-optimization-guide
- Google: AI features and your website: https://developers.google.com/search/docs/appearance/ai-features
- Generative AI performance report: https://support.google.com/webmasters/answer/16984139
- Robots meta tag, data-nosnippet and X-Robots-Tag: https://developers.google.com/search/docs/crawling-indexing/robots-meta-tag
- Pew Research Center, July 2025: https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/

---
Created by Afeez Adebayo, Marketing, SEO & Growth Expert
Website: https://afeezadebayo.com
LinkedIn: https://www.linkedin.com/in/afeez-adebayo/
X: https://x.com/za_icon
Source: https://afeezadebayo.com/skills-md/marketing/aeo-geo/ai-overviews-ai-mode/

Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). You may copy, share and adapt this skill, including for commercial use, as long as you credit Afeez Adebayo and link to the source URL above. Keep this notice in any copy or adapted version.

Workflow

Step 1. Find where AI features appear

Check the target queries in the target market and record whether an AI Overview appears, which sources it cites and whether AI Mode answers cite you.

Step 2. Read the Generative AI report

List pages with AI feature impressions. Compare with the overall Performance report to see which pages gained or lost clicks as AI impressions grew.

Step 3. Diagnose missing citations

For important queries where you rank but aren't cited, compare your page with cited pages: direct answer, specificity, evidence, freshness, visual content. For queries where you don't rank, the issue is classic SEO first.

Step 4. Cover the fan-out

Use query-fan-out-mapping to find sub-queries the overview draws on and strengthen sections that answer them.

Step 5. Review snippet controls

Check templates for unintended nosnippet or very small max-snippet values that block use in AI features. Use data-nosnippet only for text that must never be quoted.

Step 6. Decide on the opt-out

Only consider excluding the site if AI features clearly cost more than they return (for example, content licensing businesses). Model the lost impressions first.

Step 7. Improve the post-click experience

For pages cited in AI features, make sure visitors find depth, tools, data or next steps the summary couldn't give.

Images and video

Google notes AI features can include relevant images and video. Image and video SEO best practices therefore also support AI feature visibility. [Confirmed]

Local and shopping

For local businesses and retailers, Google Business Profile and Merchant Center feeds help products and services appear in AI responses. [Confirmed] See ai-commerce-product-visibility.

Worked example

A personal finance publisher's clicks for "how to calculate stamp duty" fell 40% year on year while average position stayed at 2. The Generative AI report showed the page with high AI impressions; the overview cited a government page and a competitor calculator. The publisher rebuilt its page around an interactive calculator with worked examples for first-time buyers and second homes, a direct answer passage and a dated rate table. The page became one of the cited sources, and clicks recovered part of the loss, with longer sessions and more calculator use than before.

Output format

# AI Overviews and AI Mode review: <site>, <market>, <date>
## AI feature presence for target queries (appears, cited sources, us cited?)
## Generative AI report summary (pages, impressions, trend)
## Missing-citation diagnosis by query
## Controls review (snippet controls, opt-out decision)
## Actions with owners and re-check dates

Snippet controls and their effects

ControlEffect on AI featuresUse when
nosnippetPage not used in snippets or AI featuresContent must not be quoted at all
data-nosnippet on an elementThat element excludedProtecting specific text such as prices or gated content
max-snippet:[n]Limits snippet lengthBalancing visibility and control
Search Console generative AI settingOpts the site out of AI featuresAfter modelling lost impressions

Check templates for accidental nosnippet tags; they quietly remove pages from AI Overviews.

Common mistakes

  • Blaming AI Overviews for drops caused by ranking or technical issues.
  • Using nosnippet sitewide and wondering why you're never cited.
  • Opting out without modelling lost visibility.
  • Thinking Google-Extended controls AI Overviews.
  • Expecting AI-specific clicks in Search Console; the report shows impressions.

Questions to ask before you recommend changes

  • Does the business need clicks, exposure, or both from these queries?
  • Which pages have AI feature impressions, and are clicks rising or falling there?
  • Are there legal or licensing reasons to limit snippets?

Self-check before you answer

  • Did I separate Google's documented positions from industry studies?
  • Did I use the Generative AI performance report?
  • Did I check snippet controls and explain the opt-out trade-off?
  • Are recommendations grounded in ranking for the main and fan-out queries?

Official documentation and research

When to use this skill, and when not to

Use it when AI Overviews appear for your important queries, when clicks fell while rankings held, when you want to be cited in AI Overviews or AI Mode, when deciding on snippet controls or Google’s AI opt-out, and when setting up AI feature reporting.

Google states that AI Overviews and AI Mode are rooted in its core ranking and quality systems and use retrieval and query fan-out over its index. To be eligible, a page must be indexed and snippet-eligible, and the site included in generative AI features in Search Console. So the work starts with classic SEO and adds AI-specific diagnosis and measurement.

Don’t use it for other engines (use ChatGPT Search Visibility or Perplexity, Copilot, Claude and Gemini).

Example prompts

  • “Our clicks fell 40% but rankings held. What is happening in AI Overviews and AI Mode?”
  • “Which of our pages appear in AI features? Read the Generative AI report export.”
  • “We rank second for this query but aren’t cited in the AI Overview. Why?”
  • “Should we use nosnippet or Google’s AI opt-out for our paid research?”
  • “Make our calculator page the source AI Overviews cite.”

Example output

The skill returns a fixed structure. A shortened example for an imaginary site:

# AI Overviews and AI Mode review: example.com, UK, 10 October 2026

## Presence for target queries
| Query | AI Overview? | Cited sources | Us cited? | Our rank |
| how to calculate stamp duty | Yes | gov.uk, competitor calculator | No | 2 |

## Generative AI report
/stamp-duty-calculator/: 52,000 AI feature impressions (Sep); clicks down 40% YoY.

## Diagnosis
Cited pages offer an interactive calculator and dated rate table; ours has prose only.

## Controls
No nosnippet issues. Opt-out not recommended: AI impressions outweigh licensing concerns.

Google’s documented positions are kept separate from industry click studies, so recommendations rest on the right kind of evidence.

What the agent needs from you

The skill asks for Search Console access including the Generative AI performance report, your important queries and pages by market, 16 months of click and impression trends, any snippet controls in use and how the business makes money from these queries: clicks, exposure or both.

Research behind this skill

FindingSourceLabel
Eligibility: indexed, snippet-eligible and included in generative AI features in Search ConsoleGoogle Search Central, 2026Confirmed
Opting out removes impressions and traffic from AI features but doesn’t change other rankingsGoogle, June 2026Confirmed
8% of visits with an AI summary led to a result click, versus 15% without; 1% clicked a link inside the summaryPew Research Center, July 2025Study
Brands cited in AI Overviews had higher CTR than brands not citedSeer Interactive, 2025Correlation

How this skill hands off to related skills

Google AI features need several skills. Fan-out sub-queries come from Query Fan-Out Mapping, page rewrites from Answer-Ready Content, reporting from AI Visibility Tracking and products from AI Commerce. Snippet controls follow Robots.txt and Robots Directives, and indexing problems go to the Crawlability and Indexation Audit.

How to run the Google AI Overviews and AI Mode skill step by step

  1. Install the skill as ai-overviews-ai-mode/SKILL.md.
  2. Open the Generative AI performance report in Search Console and export AI feature impressions by page.
  3. List important queries and pages by market, with 16 months of clicks and impressions.
  4. Check snippet controls in your templates: nosnippet, data-nosnippet and max-snippet.
  5. Review the diagnosis per query: AI Overview present, sources cited, you cited or not, and why.
  6. Apply fixes and decide on controls with the trade-offs the skill sets out.

Situations this skill handles

Google’s AI features change how clicks flow. The skill handles these situations:

SituationWhat the skill does
Clicks fell while rankings heldReads AI impressions and compares cited pages with yours
Ranks well but not citedAdds the direct answer, specifics and evidence cited pages have
Doesn’t rank for the queryTreats it as classic SEO first
Sitewide nosnippet in a templateFinds and removes unintended restrictions
Considering the AI opt-outModels lost impressions before recommending anything
Products and local servicesChecks Merchant Center and Business Profile data

What makes this skill different

Most AI Overview advice is speculation about secret signals. This skill starts from what Google has documented: AI Overviews and AI Mode use retrieval and query fan-out over its index, rooted in core ranking systems, with eligibility tied to indexing, snippets and a Search Console setting. It measures with Google’s own report and keeps independent click studies separate from Google’s statements.

It also focuses on the click that remains. Visitors who still click after an AI summary often want depth, tools or data, so cited pages should deliver what the summary can’t.

Practical tips from real projects

  • Google-Extended doesn’t control AI Overviews; the Search Console setting does.
  • The report shows impressions, not AI-only clicks.
  • Images and video help; Google says AI features can include them.
  • Opt out only after modelling the lost visibility.

A worked example: a finance publisher’s clicks for a stamp duty query fell 40% while it stayed at position two. The Generative AI report showed high AI impressions; the overview cited a government page and a competitor’s calculator. The publisher rebuilt its page around an interactive calculator with worked examples and a dated rate table. It became a cited source, recovered part of the lost clicks and saw longer sessions than before.

Independent studies put the scale in context: Pew found people clicked a result on 8% of searches with an AI summary versus 15% without. Being cited and useful is the practical response.

Frequently asked questions

Rank well for the main query or the sub-queries Google runs, have a clear, specific and original answer on the page, and make sure the page is indexed, snippet-eligible and included in generative AI features in Search Console.

No. Google-Extended controls some Gemini uses. Google Search and AI Overviews use Googlebot, and the Search Console setting controls AI feature inclusion.

You can see impressions in the Generative AI performance report. Clicks from AI features are included in the overall Performance report but not separated.

Rarely. Opting out removes all impressions and traffic from AI features. It may suit businesses whose content is the product, but model the loss first.

Yes. nosnippet, data-nosnippet and max-snippet limit what Google can show, including in AI features. Check templates for unintended restrictions.

AI summaries answer many informational questions on the results page. Independent studies show lower click rates when they appear. Being cited and offering depth helps recover some clicks.

Changelog

  • 1.0.0, 10 October 2026: first release.

Use, adapt and share this skill under CC BY 4.0: credit Afeez Adebayo and link to this page.