Quick answer: Google Search Console now has a Search Generative AI performance report (rolled out to all sites by August 31, 2026) that shows how often your pages appear in AI Overviews and AI Mode, but it does not show clicks or the queries involved. To measure traffic loss you compare clicks and click-through rate for queries whose average position has not changed, before and after the AI Overview started appearing, then subtract seasonality and any ranking change. For most home service websites the loss is concentrated in informational searches such as ‘why is my AC not cooling’, while ‘AC repair near me’ searches are largely untouched.
If your plumbing, HVAC, roofing or landscaping website has lost organic clicks this year while your rankings look fine, there is a good chance an AI Overview is answering the question before anyone reaches your link. The problem for owners is proving it. Google folds AI clicks into ordinary web search totals, so the loss never shows up as a line item.
This guide walks through exactly what Search Console reports in late 2026, the six-step method Everyday Media Group uses to put a number on AI Overview loss for home service clients across Dallas-Fort Worth, and how to decide which of those lost clicks were ever worth having. You will not need a paid rank tracker to follow it.
What does Google Search Console actually show about AI Overviews in 2026?
Search Console shows AI impressions, not AI clicks. On June 3, 2026 Google announced a dedicated Search Generative AI performance report, and by August 31, 2026 it had reached every verified property worldwide. The report tells you how many times links to your site appeared inside AI Overviews and AI Mode, broken down by page, country, device and date.
What it does not tell you is just as important for measurement:
- No clicks. Clicks from an AI Overview link are still counted in the regular Performance report under the Web search type, mixed in with classic blue-link clicks.
- No queries. The AI report does not list which searches triggered the appearance.
- No separation. AI Overviews and AI Mode are reported together.
- No API. As of this writing the data lives only in the Search Console interface, so it cannot be pulled into Looker Studio or BigQuery automatically.
- No Search Appearance filter. Several articles still claim there is an AI Overviews option under Search Appearance. Google’s own documentation confirms there is not.
One more thing changed in the same update: a control in Search Console that lets you opt your site out of AI Overviews and AI Mode without affecting normal rankings. We rarely recommend using it for a local service business, and the reason is covered later in this post.
The practical takeaway is that AI loss has to be inferred from the relationship between impressions, clicks, click-through rate and position in the standard report. That is what the rest of this method does.
Why are AI Overviews hitting home service websites in particular?
Because a large share of home service search demand is diagnostic, and diagnostic questions are exactly what AI Overviews answer. A homeowner in Southlake typing ‘why does my water heater make a popping noise’ wants a cause, not a company. Google’s AI answer gives the cause (sediment buildup) in two sentences, so the homeowner never scrolls to the blog post that used to rank first.
Across the home service accounts we manage in Dallas-Fort Worth, the pattern is consistent. Posts built around ‘why’, ‘how to fix’, ‘what causes’, ‘is it normal’ and ‘how long does it take’ lose the most. Service and city pages aimed at ‘repair near me’, ‘emergency plumber Plano’ or ‘roof replacement Fort Worth’ hold steady, because the searcher wants a provider and Google still shows the local pack and the organic listings for those.
That split matters for measurement. If you look at sitewide clicks you will see a modest decline and conclude everything is slipping. If you look at the blog separately from the money pages, you will usually see a sharp blog decline and flat or rising service-page traffic. The second picture is the true one, and it changes what you do next. We covered the local-search side of this shift in how AI Overviews change local search in DFW.
How do you measure AI Overview traffic loss in Search Console step by step?
Build a query-level baseline, isolate queries where position held but click-through rate fell, then quantify the click gap against that baseline. Here is the six-step version we run for clients. Budget about two hours the first time.
- Export a long baseline. In Performance, choose Search results, set the date range to the last 16 months (the maximum), switch on Clicks, Impressions, CTR and Position, and export the Queries table and the Pages table to a spreadsheet. The interface only holds 16 months, so keep this export outside Search Console as your permanent baseline.
- Mark the AI start date for your market. AI Overviews reached most United States searches in May 2024 and expanded through 2025 and 2026. For each query cluster, note the month impressions rose or stayed flat while clicks dropped. That is your before and after boundary for that cluster.
- Flag the ‘position held, CTR fell’ queries. Add a column that compares average position before and after. Keep only queries where position moved less than one place. Among those, flag every query where CTR fell by 25 percent or more. These are your AI Overview suspects, because nothing else in the listing explains a click drop at a stable rank.
- Confirm with the Generative AI report. Open the new report and check the Pages tab. If the landing pages for your flagged queries show AI impressions, the pattern is confirmed from the other direction: Google is citing the page in an answer, which is where the clicks went.
- Quantify the gap. For each flagged query, multiply current monthly impressions by the old CTR to get expected clicks, then subtract actual clicks. Sum the result. That figure is your estimated monthly click loss to AI Overviews.
- Segment by intent before you react. Tag each flagged query as diagnostic, how-to, comparison or provider-seeking. Then pull your call tracking or form data for the landing pages. A diagnostic post that lost 400 clicks and never generated a call is a different problem from a comparison page that lost 60 clicks and used to produce three jobs a month.
A worked example from a typical DFW plumbing site: a ‘water heater popping noise’ post held position 2.4 from March 2025 to March 2026, impressions rose from 3,100 to 3,600 per month, and clicks fell from 410 to 140. Expected clicks at the old CTR would be about 475, so the estimated monthly loss is roughly 335 clicks. The water heater repair service page on the same site gained 11 percent over the same window.
How do you tell AI Overview loss apart from seasonality, a core update or a ranking drop?
Run the confounder checklist before you blame AI, because three other causes produce similar charts. Each one has a signature that the baseline export exposes.
| Cause | Impressions | Clicks | Position | How to confirm |
|---|---|---|---|---|
| AI Overview answering the query | Flat or rising | Down sharply | Stable | Page shows in the Generative AI report; query is a question |
| Seasonality | Down | Down in proportion | Stable | Same dip in the prior year’s export for the same months |
| Ranking loss (core update or competitor) | Down | Down | Worse by 2 or more | Position change clusters around a known update date |
| Tracking or canonical issue | Split across two URLs | Split | Looks unstable | Pages report shows a UTM or duplicate URL earning its own impressions |
The last row is more common than owners expect. Many home service sites have a Google Business Profile link with a tracking parameter that gets indexed as its own URL, so clicks split between two versions of the homepage. That looks like a loss on the clean URL and is fixed with a canonical tag, not with content changes. A full SEO audit catches this alongside the AI analysis.
If you want an external check, Google’s own Performance report documentation explains how impressions and positions are counted, including the rule that a page appearing in both an AI Overview and the classic results counts as one impression at the higher position.
What should a home service company do with the lost queries?
Keep informational content, restructure it to be cited, and stop measuring it by clicks alone. Opting out of AI features or deleting the blog are the two most common overreactions, and both usually cost more than the clicks that were lost.
Queries that never converted
If the landing page produced no calls or forms over 12 months, the lost clicks were vanity traffic. Leave the post, add a clear answer block at the top so Google can cite it, and move on. Being the cited source still builds brand recall in your service area even when the click does not happen.
Queries that did convert
These deserve work. Rewrite the opening so the direct answer appears in the first 60 words, use question-format subheadings that match the search, add a concise FAQ with FAQPage schema, and include a next-step sentence that names your service and city. AI Overviews cite sources that answer plainly, and the structured answer makes the page more likely to be the one named. Our answer engine optimization guide for DFW businesses goes deeper on that format.
Provider-seeking queries that slipped
If ‘AC repair near me’ style queries lost ground, AI is probably not the cause. Check the local pack, Google Business Profile categories, review velocity and the service page itself. Those are local SEO problems with well-known fixes.
Why we do not recommend the opt-out toggle
Google’s AI features guidance is clear that AI Overviews and AI Mode draw on the same index as regular search. Opting out removes your pages from the answer that most homeowners now read first. For a local company whose competitors stay in, that is a visibility loss with no offsetting gain.
How should you report AI Overview impact going forward?
Track three numbers monthly and report them beside each other. First, AI impressions from the Generative AI report, which shows whether Google is citing you more or less. Second, the estimated click loss from the six-step method, recalculated for the flagged query set. Third, calls and forms from the affected landing pages, which is the number the business actually feels.
Because the AI report cannot be exported through the API yet, note the monthly impression total by hand in the same spreadsheet as your baseline. When Google adds API access, the history will already be there. Keep the Web search type Performance export running monthly too, since that is where any AI clicks you do earn will continue to land.
Google first described this direction when it expanded AI Overviews and introduced AI Mode in March 2025, and the reporting has changed several times since. Whatever method you adopt, date it, so a future drop can be read against the rules that applied at the time.
Want the AI Overview loss analysis done for your home service website?
Everyday Media Group is a Southlake, TX digital marketing agency that runs this exact measurement for plumbing, HVAC, roofing, irrigation and remodeling companies across Dallas-Fort Worth and beyond. Our AI search optimization services include the baseline export, the flagged-query analysis, the confounder check and a prioritized rewrite plan for the pages that still matter to your phone.
Call 817-286-7730 or contact Everyday Media Group to request a Search Console review. Share read-only access to your property and we will return the estimated monthly loss, the queries behind it and the fixes in priority order.
Frequently asked questions about measuring AI Overview traffic loss
Does Google Search Console show clicks from AI Overviews?
No. As of October 2026 the Search Generative AI performance report shows impressions only. Clicks from AI Overview links are counted inside the regular Performance report under the Web search type, with no filter to separate them from classic organic clicks.
Is there a Search Appearance filter for AI Overviews in Search Console?
No. A claim that such a filter existed circulated in 2025 and still appears in some articles, but Google’s documentation confirms AI Overviews and AI Mode are reported under the Web search type without a Search Appearance option.
Why do I not see the Generative AI report in my Search Console?
The report only appears for properties whose pages have already been shown in AI Overviews or AI Mode. If your site has not been cited yet, the report stays hidden. It rolled out to all eligible sites worldwide on August 31, 2026.
What is the clearest sign that AI Overviews are taking my clicks?
A query where average position stayed the same, impressions held or rose, and click-through rate fell by a quarter or more. That combination means people saw your listing as often as before but had less reason to click, which is what an AI answer above the results produces.
Which home service searches lose the most traffic to AI Overviews?
Diagnostic and how-to searches lose the most: why an AC is not cooling, what causes a sprinkler head to leak, how long a kitchen remodel takes. Provider-seeking searches such as ‘plumber near me’ or ‘roofing company Fort Worth’ are largely unaffected because the searcher wants a business, not an explanation.
How many months of data do I need to measure AI Overview loss?
Export the full 16 months Search Console allows. You need at least a year so that the same months can be compared against the prior year, which is how seasonality is separated from AI loss.
Can Google Analytics 4 show AI Overview traffic?
Not for Google’s own AI features. Those sessions arrive as google / organic like any other search click. GA4 can show referral sessions from non-Google AI tools such as ChatGPT or Perplexity, which appear as referrals from those domains.
Should I opt my website out of AI Overviews using the new Search Console control?
For a local service business, usually not. Opting out removes your pages from the answer most homeowners read first while your competitors remain in it, and the clicks that the AI answer replaced were often low-converting informational visits anyway.
Can I pull the AI Overview impression data into Looker Studio or BigQuery?
Not yet. The Generative AI performance report is available only in the Search Console interface and has no API endpoint as of October 2026, so monthly totals have to be recorded by hand until Google adds programmatic access.
How does Everyday Media Group measure AI Overview impact for clients?
We export the 16-month query and page baselines, flag queries where position held but CTR fell, confirm against the Generative AI report, calculate expected versus actual clicks, segment by intent, and match the result against call tracking and form data so the loss is expressed in leads as well as clicks.