AI Overviews Organic Traffic Loss: Search Console Audit
If your impressions are flat and your clicks are falling, AI Overviews are the first suspect. Search Console will show you which queries and pages are affected within an afternoon. The fix is to fight for top-three rankings on the queries where an AI Overview appears and to measure revenue instead of raw clicks. Don't chase "AI Overview optimization" tricks. In our own data, being cited in an AI Overview tracks organic rank closely. Rank is still the lever you control.
The number circulating right now is stark. Reporting out of France puts the click loss at up to 69% on affected queries, while impressions held steady. We haven't validated that figure against our own client data, so treat it as a ceiling on what's possible. Don't read it as a forecast for your site. The pattern behind it is what matters: Google still shows your listing, and fewer people click it because the answer sits above you on the results page.
Here's how to confirm whether that's happening to you, how to separate it from the other causes of a traffic drop, and what to change once you know.
The signature of AI Overviews organic traffic loss
AI Overviews organic traffic loss has a specific shape in Search Console. Four metrics move together:
- Impressions: flat or rising
- Average position: flat
- Clicks: falling
- CTR: falling, often sharply
That combination is the tell. Your page still ranks where it did and still gets seen, but searchers get what they needed from the summary and never reach you.
Compare that with the other common patterns:
| Pattern | Impressions | Position | CTR | Likely cause |
|---|---|---|---|---|
| AI Overview absorption | Flat/up | Flat | Down | Answer shown above your listing |
| Ranking loss | Down | Down | Down/flat | Algorithm update, competitor gains, technical issue |
| Demand drop | Down | Flat | Flat | Seasonality, fewer people searching |
| Indexing problem | Down sharply | N/A for affected pages | N/A | Pages dropped from the index |
If your position dropped, you have a ranking problem. Putting AI Overviews in the headline won't change the diagnosis. Work through our guide to the November 2026 Google core update before you touch anything else. If impressions collapsed without a position change, look at demand or indexing.
Before you trust any of these numbers, make sure the data itself is clean. Inflated impressions from junk or spam queries can make a CTR drop look worse than it is. Read Is Your Search Console Data Real? first if you see impressions spiking on queries that have nothing to do with your business.
Step 1: Set up a clean comparison in Search Console
Open Performance > Search results and set a date comparison.
- Pick comparable windows. Year over year is cleanest for seasonal businesses. For a store that sells patio furniture, comparing October to September tells you about the season and nothing about AI Overviews. Use the last three months against the same three months a year ago. If your business has no seasonality, three months against the prior three months works.
- Turn on all four metrics: clicks, impressions, CTR, and average position.
- Filter to Web search type. Image and video searches behave differently and muddy the picture.
- Export the Queries tab and the Pages tab separately. The interface shows only the top 1,000 rows. If you have more queries than that, pull the data through the Search Console API or a connected reporting tool so the long tail doesn't disappear.
In your spreadsheet, add four columns per row: click change, impression change, CTR change, and position change. Express each as a percentage, except position, which you'll want as an absolute difference.
Step 2: Flag the queries that match the AI Overview signature
Now filter for the signature. Set a tolerance band that suits your data volume. As a starting point, use impressions within ±10% of the prior period, position within one spot, and clicks down by more than 20%. Tighten or loosen the band depending on how noisy your numbers are. Low-volume queries swing wildly, so set a minimum impressions floor before you trust any individual row.
What's left is your candidate list: queries where you held your ground on visibility and lost the click anyway.
Then confirm the AI Overview exists. At the time of writing, Search Console's standard Performance report doesn't give you a filter that separates AI Overview appearances from regular results. Your candidate list is a list of suspects, not a confirmed list. Verify it in one of two ways:
- Manual checks. Search your top 20 to 30 candidate queries in a clean browser session, ideally from the location your customers search from. Note whether an AI Overview appears, whether it cites you, and which page it cites.
- Rank tracking with SERP feature data. If you track rankings, check whether your tool records AI Overview presence per keyword. This scales better than manual checks and gives you history.
Record three things per query: AI Overview present (yes/no), your site cited (yes/no), and the cited URL if it's yours.
Step 3: Segment by query type
All lost clicks are not equal. Run three regex filters in the Queries report. Search Console accepts RE2 syntax under Query > Custom (regex).
Brand vs. non-brand. Filter queries matching your brand name and its common misspellings, for example (? i)yourbrand|your brand|yourbrnd. Then filter for the inverse. Branded queries rarely lose much to AI Overviews because the searcher wants you specifically. If your branded clicks are dropping, look somewhere else first: reputation, a competitor bidding on your name, or a navigation problem.
Question queries. Filter for queries that start with a question word:
^(what|how|why|when|which|who|is|are|can|does|do|should)\b
These are the queries where AI Overviews appear most often and do the most damage. If the bulk of your click loss sits here, you're looking at informational content losing traffic that rarely converted directly.
Comparison queries. Filter for comparison language:
(? i)\b(vs|versus|compare|comparison|best|alternative|alternatives|or)\b
Watch this segment closely. Our own data shows comparison queries behave differently, which we'll get to below.
Once segmented, you'll usually find the loss concentrated in one or two segments. That concentration tells you where to spend effort.
Step 4: Move from queries to pages
Queries tell you what's happening. Pages tell you what to fix.
Export the Pages tab with the same date comparison and group URLs by template. On an e-commerce site, that typically means:
- Blog and guide content
- Category pages
- Product pages
- Brand or collection pages
- Support and FAQ content
Use page filters with regex to isolate templates, for example /blog/|/journal/|/guides/ for content and /category/|/collections/ for category pages, adjusted to your URL structure.
Then click into the pages losing the most clicks and open the Queries tab for that page. You'll see which queries drove the loss. A buying guide might hold its position on 40 queries and lose clicks on the six that now trigger an AI Overview. That's a much narrower problem than "our blog traffic is down."
Step 5: Connect the loss to revenue before you react
This is where a lot of panic gets resolved. Open GA4 and pull conversions and revenue for organic landing pages over the same windows.
You'll typically find one of two situations:
Situation A: Clicks down, revenue flat. The lost clicks were people who wanted a quick answer and were never going to buy. An AI Overview answering "how long does a car wrap last" takes a visitor who would have read a paragraph and left. That's a traffic loss you can live with. Don't rebuild your content strategy around winning those clicks back.
Situation B: Clicks down, revenue down. The lost clicks included buyers. This is the urgent case, and it usually shows up on commercial and comparison queries, or on guides that fed visitors into product pages.
Check your attribution while you're in GA4. Some AI-referred visits land in Direct instead of Organic, which can make an organic decline look bigger than the real change in demand. If your Direct traffic rose while organic fell, investigate before you conclude anything.
What our own data says about AI Overview citations
Once you know which queries trigger AI Overviews, the next question is how to get cited in them. We ran a pooled analysis on this across client sites we manage, from July through September 2026: 2,350 tracked keyword snapshots where an AI Overview appeared, covering 314 keywords.
How often the AI Overview cited the site, by the site's organic position:
| Organic position | Cited in the AI Overview |
|---|---|
| 1 | 73% |
| 2–3 | 49% |
| 4–10 | 42% |
| 11–20 | 9% |
| Beyond 20 or not ranking | 3% |
The pattern held on nearly 80% of the sites in the analysis. The rest had too little data to judge. We wrote this up in more detail in AI Overview Citations Follow Rank.
Two more findings change how you should act on your audit:
- About a quarter of citations went to a different page on the same site than the page that ranked. Google ranked one URL and cited another. If you only check whether your ranking page got cited, you'll miss a large share of your citations.
- Citations without a top-20 rank were mostly comparison queries. Comparison content is the one place where a lower-ranking page has a real shot at the summary.
Here's where we could be wrong. This is a correlation, and we haven't established causation. It's possible that whatever makes a page rank first also makes it citable, and rank itself does none of the work. That wouldn't change the practical advice much. The work that improves rank is the same work that appears to improve citation. If Google changed how it selects citations and the gap between position 1 and position 10 closed, we'd revise this.
How to adapt your SEO once you have the audit
1. Push your 4–10 rankings into the top three on AI Overview queries
Look at the queries from Step 2 where an AI Overview appears and you rank 4 through 10. In our data, citation rates climb from 42% to 49% to 73% as you move from page one into the top three and then to the top spot. These are your highest-leverage targets.
The work is ordinary SEO done thoroughly: better coverage of what the searcher needs, stronger internal links pointing at the page, cleaner technical health, and better answers than the pages above you. There's no separate AI Overview playbook that skips this.
Spend less on queries where you rank 11 to 20 with no comparison intent. A 9% citation rate means that a lift from 15th to 12th position doesn't do much for AI Overview visibility. Either commit to a page-one push on those queries or move the budget elsewhere.
2. Make the cited page the page you want cited
Since about a quarter of citations land on a different page than the one ranking, check which URL Google cites for your priority queries. You'll find cases where a three-year-old blog post gets cited while your category page holds the ranking. Sometimes it's the reverse.
Decide which page should own the query. Then:
- Consolidate overlapping pages so two URLs aren't competing for the same answer. Redirect the weaker one or merge the content.
- Point internal links with descriptive anchor text at the page you want to own the topic.
- Put the direct answer near the top of that page, in plain language, followed by the depth that earns the click.
Internal linking problems run deeper than most owners expect. In one technical SEO audit, we found crawl issues on 12,000 orphaned product URLs, pages that no internal link pointed to. Pages Google struggles to reach are pages it struggles to rank or cite. Run a crawl and compare it against your sitemap and your Search Console indexed pages before you write a single new paragraph.
3. Build comparison content on purpose
Comparison queries are the exception in our citation data, so treat them as their own category. If your audit shows comparison queries losing clicks or you have no comparison content at all, build it:
- Product vs. product pages for the items your customers actually weigh against each other
- "Best X for Y" guides tied to real use cases you sell into
- Alternative pages for the products or approaches buyers are moving away from
Be specific. Use tables, real specs, and clear recommendations for who should buy what. A comparison page that refuses to recommend anything gives the searcher no reason to click past the summary.
4. Let informational traffic go where it never converted
If Step 5 showed Situation A, informational clicks dropping with revenue flat, don't rewrite every guide in an effort to win those clicks back. Those visitors were getting a fast answer, and now Google gives it to them directly.
Keep the content. It still earns impressions, supports topical authority, and can still get cited. But move your effort toward pages that sit closer to a purchase decision. Update the guides that feed buyers into category and product pages. Leave the trivia alone.
5. Strengthen the pages AI Overviews can't replace
An AI Overview can explain what a 32 oz insulated bottle is. It can't let someone pick a color, check stock, and check out. Product and category pages carry intent that a summary doesn't satisfy, and they deserve more of your attention as informational clicks decline:
- Clean structured data on products: price, availability, reviews, and shipping where applicable
- Category pages with real filtering and useful copy above the grid, instead of a block of keyword text below it
- Fast, stable pages, especially through cart and checkout. We once inherited a store whose checkout took 11 seconds to load because nobody had touched the database indexing in three years. Ranking gains don't help much if the buyer leaves before the page renders.
This is where SEO and development have to work together. A category page that needs better internal linking, a faster template, and new structured data is three tickets for three different vendors at a lot of agencies. When one team owns all three, the fix ships in one pass.
6. Change what you report
If your monthly report leads with organic clicks, it will tell you you're failing even when revenue holds. Rebuild your reporting around:
- Organic revenue and conversions by landing page template
- Click and CTR trends on queries where AI Overviews appear, tracked separately from the rest
- AI Overview citation status for your priority queries, including which URL got cited
- Position on your top-three push list
That gives you a view of whether you're winning the clicks that pay for themselves, which is the only question that matters.
A repeatable audit schedule
Make this a quarterly routine instead of a one-time exercise. AI Overviews appear on new queries and drop off others, and your candidate list will shift.
Quarterly:
- Rerun the year-over-year comparison and flag signature queries.
- Re-verify AI Overview presence and citations on your top candidates.
- Update the segment split: brand, question, and comparison.
- Check revenue against clicks by template.
- Reprioritize the top-three push list.
Monthly:
- Track position on the push list.
- Spot-check citation status on your 10 most valuable AI Overview queries.
After any major Google update:
- Rerun the full audit. A core update can turn an AI Overview problem into a ranking problem overnight, and the fixes are different.
The bottom line
Flat impressions with falling clicks means AI Overviews are taking the answer above your listing. Search Console will show you exactly which queries and pages are affected if you compare the right windows, segment by query type, and verify the SERP by hand. From there, the response is focused: push your page-one rankings into the top three on AI Overview queries, make sure the right page gets cited, build comparison content, and judge the result by revenue.
That combination of technical SEO, content, and development work is what our SEO and organic growth services are built around: one team running the audit, writing the content, and shipping the template fixes. If you want us to run this Search Console audit on your site and hand you a prioritized list of the queries and pages worth fighting for, get an audit. If you already know you need a plan for the next year of search, request a proposal.