AI Overview ads: how Google places ads in AI answers and how a B2B team should test them
Ads inside Google AI Overviews reach you through your match types, report in aggregate and have no off switch, so a B2B team tests them through contextual matching and judges them on cost per opportunity.

AI Overview ads are Google Search and Shopping ads, labeled "Sponsored", that appear inside, above or below the AI-generated answer on a results page. As of November 2026, ads inside the answer run in English in the US and 11 other countries. Existing Search, Shopping and Performance Max campaigns join automatically, with no separate opt-out.
You've probably had the question already. Someone in a leadership meeting saw a "Sponsored" block inside an AI answer, and the message that followed was some version of "are we in there, and if not, why not?" You open the account to check and find that there's no AI Overviews column, no segment and no setting with that name.
That's because Google reports the placement inside your normal numbers, and the one lever it does give you sits in your match types. This article is for the head of paid media or PPC lead who has to answer the question properly. It covers where the ads sit, what makes yours eligible, what the reporting shows, and how to test the placement on pipeline when there's no switch to flip. Scalewell helps paid teams answer questions like this in their weekly reporting, so the decision about AI Overview placements rests on pipeline rather than clicks.
Ads appear in three positions around an AI Overview
Google ads can show above an AI Overview, below it, or within it, and the help page on ads in AI-powered experiences in Google Search says all ads inside AI Overviews and AI Mode are "clearly labeled as 'Sponsored'". Ads above and below the answer are ordinary search ads that happen to sit next to an AI Overview, and those are available everywhere AI Overviews are, which Google puts at "All 200+ countries and territories, and all languages".
The placement inside the answer is the new inventory. The same help page says your ad "can appear both above an AI Overview and within it on the same search", but ads "don't appear both within an AI Overview and immediately below it". Text ads and Shopping ads are the two formats that serve there today, and Google says it is "also experimenting with other types of ads such as local inventory ads and vehicle ads".
Some placements also carry an AI explainer, which is short text written by Google's Gemini models to say how an ad relates to the question. Advertisers "can't manually write, edit, approve, or opt out of them", and the help page says they draw on "publicly available sources including your website".

The rollout took two years. Google announced on October 3, 2024 that "Ads in AI Overviews are now available for mobile users in the US", extended them to desktop at Google Marketing Live on May 21, 2025, and the help page now lists 12 countries. The audience is large: Sundar Pichai said at Google I/O on May 19, 2026 that "AI Overviews now has over 2.5 billion monthly active users."
Search, Shopping and Performance Max campaigns are eligible by default
Existing Search, Shopping and Performance Max campaigns are "automatically eligible to serve ads within AI Overviews and AI Mode", according to Google's help page, and they need no new campaign or separate budget. Billing works the same way too: the help page says these ads use "the same auction mechanism and cost-per-click (CPC) or conversion-based bidding models" as standard search ads, and "You aren't charged extra to participate."
Where the ads can serve inside the answer is narrower than where AI Overviews appear. Here's the list as Google publishes it in November 2026.
| Placement | Countries | Languages | Notes |
|---|---|---|---|
| Above or below an AI Overview | Everywhere AI Overviews are available | All | Standard search ads |
| Inside an AI Overview | United States, Australia, Canada, India, Indonesia, Kenya, Malaysia, New Zealand, Nigeria, Pakistan, Philippines, Singapore | English | Indonesia covers AI Overviews only; Pakistan text ads only |
There's no UK, no EU country and no non-English language on the list, so a team running Search in Germany today buys the positions above and below the answer and nothing inside it. Google says the list "may not be exhaustive", so check the page before you plan a non-US test.
Your match types decide whether you show on research questions
Google matches ads in AI Overviews in two ways, and which one your campaign can use depends on how it targets. The help page describes direct matching "to the user's query (when the search already expresses clear commercial intent)" and contextual matching "to both the user's query and the AI response" when someone asks "a longer, exploratory, or conversational question" and the commercial intent shows up in the answer's next steps.
This is the rule that matters most for a B2B account: exact and phrase match keywords "aren't eligible for contextual matching on broader exploratory questions", while broad match, AI Max for Search, Shopping and Performance Max can use both kinds of matching.
| Targeting | Direct query matching | Contextual matching (query plus AI answer) |
|---|---|---|
| Exact match keywords | Yes, when the query expresses direct commercial intent | No |
| Phrase match keywords | Yes, when the query expresses direct commercial intent | No |
| Broad match keywords | Yes | Yes |
| AI Max for Search (search term matching on) | Yes | Yes |
| Performance Max | Yes | Yes |
| Shopping campaigns | Yes | Yes |
Now put that next to how B2B buyers search. Seer Interactive's April 2026 study of 53 brands and 5.47 million tracked queries found that informational queries showed an AI Overview 36% of the time, against 8% for commercial queries and 5% for transactional ones. Among informational queries, comparison searches triggered one 95.4% of the time and question-format searches 85.9% of the time.
Those are the "X vs Y" and "how do I" searches a buying committee runs months before it fills in a demo form, and an exact and phrase match account can't reach them through contextual matching. So a tightly built B2B account, the kind that took years of negative keyword work to get clean, is mostly absent from the placement inside the answer on research queries by design. Treat that as a trade you now make knowingly: stay tight and miss the research-stage placement, or loosen matching and pay for a wider and messier set of questions.
If you're weighing that trade, our post on AI Max for Google Ads covers the controls and the experiment setup for search term matching in detail.
You can't opt out, and the reporting won't separate the placement
Google's help page answers the opt-out question directly: "you can't target ads exclusively to AI Overviews or AI Mode, or opt out of showing ads in them separately from Search." The controls you keep are the ones you already have, and Google names "negative keywords, asset pinning" and brand exclusions.
Reporting is aggregated: impressions, clicks, conversions and cost from AI Overviews and AI Mode sit inside your normal campaign reports, and the queries that triggered them sit inside the search terms report "in aggregate", per the same page. Google's help pages document no segment, column or filter that splits the placement out as of November 2026, so any view of it has to come from a proxy.
The search terms themselves are also less literal for these placements. Google's page on ad group and asset group prioritization says that for "searches on Lens, AI Mode, AI Overviews, or auto-complete searches", "the search term shown in your reporting represents the best approximation of the user's intent." In practice, some rows in your report are Google's summary of a conversational question rather than the words typed. That matters when you build negatives from them, which our post on reviewing search terms for negative keywords every week goes into.
The independent data covers AI Overviews on the page, not ads inside them
The best public data measures what happens to paid clicks when an AI Overview appears on the results page, and no published study we found isolates ads inside the answer.
Seer Interactive's April 2026 study found paid CTR on queries with an AI Overview rose from 14.6% at the start of 2025 to 16.2% in February 2026. On queries without one, paid CTR fell from 26% to 21.8%. Seer's reading is that "Paid is relatively insulated from AIO organic effects", and that paid CTR was the most predictable series in the dataset.
What does that mean for you? Searches with an AI Overview still click through on ads at a lower rate than searches without one, but the gap narrowed through 2025, and you can forecast paid on these queries with more confidence than organic. It says nothing about cost per opportunity from the placement inside the answer, and we found no credible public figure on that for B2B or anyone else, so you'll need your own.
Keep the channel in proportion, too. For most B2B teams, LinkedIn is where you choose the buyer by job title and company, and Google is where you meet the buyer who's already asking. AI Overview ads stretch the Google job into earlier, vaguer questions without touching the LinkedIn one, and our guide to B2B paid media covers how the two budgets fit together.
Four checks show whether your ads can serve in AI Overviews
You can't see the placement directly, so you build the picture from four checks, each ending in a fact you can write down.
- Confirm geography and language. List every campaign targeting a country on Google's eligible list with English as a targeted language. Only those campaigns can serve inside the answer. Everything else can only appear above or below it.
- Map targeting to matching. For each eligible campaign, note whether it runs broad match, AI Max search term matching, Performance Max or Shopping, or only exact and phrase. The first group can match contextually; the second can't. You end with a short list of campaigns that can reach exploratory questions.
- Pull the long questions from the search terms report. Export 90 days of search terms for those campaigns and filter for rows of eight words or more, or rows that start with how, what, which, why, can or should. These are a proxy for the conversational queries where AI Overviews appear, and you're looking for how much spend and how many conversions they carry. Treat it as an estimate, since Google approximates some of these terms and doesn't label the placement.
- Spot-check the results page. Search a dozen of your highest-spend question terms from an eligible country and note when an AI Overview appears and whose ads sit inside it. The layouts vary by user and change often, so this tells you about competitors and formats, never about your share.
The result is a one-page note on which campaigns are eligible, what share of their spend goes to question-style queries, and which competitors you saw inside the answers.
Test contextual matching, because the placement has no switch
Contextual matching is the setting to test, because it decides whether a campaign can reach the exploratory questions inside the answer. In practice that means running a Google Ads experiment where the treatment adds broad match or AI Max search term matching to a campaign that currently runs on exact and phrase, in an eligible country.
- Pick a research-theme campaign. Choose a campaign whose topic generates comparison and how-to searches, such as a category term rather than your brand. Brand searches already express clear commercial intent, so exact and phrase match serve on them directly and the test would tell you little.
- Fix the conversion first. Bid toward a conversion sales recognizes, ideally an offline conversion import of qualified opportunities. Google's own best practice here asks for "accurate first-party conversion measurement", and contextual matching without it will optimize toward whatever form fill is cheapest.
- Write the exclusions before launch. Add brand exclusions and the negatives you'd want on any broad match campaign, such as jobs, free, template and course terms.
- Run a 50/50 experiment for long enough to see opportunities. If your sales cycle runs about 60 days, that means serving long enough to collect leads and then roughly 60 more days for them to become opportunities. Don't call it on week-two conversion counts.
- Compare the cohorts in your CRM. Split treatment results into question-style terms (the same proxy as in the checks above) and the rest, and compare cost per qualified opportunity for each arm.
As a hypothetical, imagine Company A's treatment arm costs 22% more and produces 30% more form fills. Most of the new fills come from question-style terms, and the CRM shows those leads becoming opportunities at half the rate of the control arm's. On cost per opportunity, the treatment loses, even though the experiment summary in Google Ads shows it winning. Company B runs the same test and finds the question-style cohort converting to opportunity at a similar rate to the control. Company B keeps the broader matching, Company A reverts, and both are right.
The rule we'd use: keep contextual matching only if cost per qualified opportunity in the treatment arm is no worse than control within the sales cycle you measured, and the share of spend on queries sales would never call stays below a ceiling you set before the test started.
Research traffic, explainers and leaky negatives cause most problems
None of the three problems below shows in the experiment summary.
The first is research-stage traffic that never buys. Contextual matching reaches people asking how a category works, which for many B2B products includes students, job seekers and competitors, and sales notices before the dashboard does.
The second is the AI explainer. It can describe your product using whatever is on your site, so the copy beside your ad may include a sentence nobody on your team wrote. If your pricing page is two releases out of date, the explainer may repeat it.
The third is leaky negatives. A negative built on one approximated search term may not block the next phrasing of the same question, so expect to review question-style terms more often in the first month.
If your accounts run outside the eligible countries or in a language other than English, skip the test for now, because there's no placement inside the answer for you to reach yet.
AI Mode ads are a separate placement with their own rollout
AI Mode is Google's conversational search tab, and ads there follow the same eligibility rules but serve across follow-up questions. Google announced on October 28, 2026 that it is "launching multiturn ads in the U.S." for follow-up queries and extending AI Mode ads to ten more countries, with Brazil, Mexico and the UK planned for Q4 2026. Our guide to AI Mode ads covers how those differ, and our overview of AI search advertising compares Google's placements with ads in other AI answers.
Where to start
Run the four checks above this week and write down which of your campaigns can match contextually at all. If none can, you've learned that the placement isn't part of your account yet, and the experiment is the next decision. Scalewell helps paid teams set up tests like this one and read them against pipeline.
Frequently asked questions
Are AI Overview ads available in the UK or Europe?
Not inside the answer, as of November 2026. Google's help page lists 12 countries for ads within AI Overviews, all English-language, and no UK or EU country is among them. Standard ads still appear above and below AI Overviews in every country where AI Overviews run. Google has said AI Mode ads are planned for the UK in Q4 2026.
Do ads in AI Overviews cost more than normal search ads?
No. Google says ads in AI Overviews and AI Mode run in the same auction and use the same cost-per-click or conversion-based bidding as standard search ads, with no extra charge to take part. What can change is your average cost, because contextual matching brings in a different mix of queries with different competition.
Does buying Google ads help my brand get cited in AI Overviews?
No. Google states that organic AI Overview and AI Mode answers are generated independently and that buying ads doesn't influence them. It also says whether your site is cited organically doesn't affect your ad auction eligibility.
How do I keep a campaign out of AI Overview ads?
You can't fully, because Google offers no opt-out from the placement separate from Search. The closest control is targeting: a campaign built only on exact and phrase match keywords can still serve inside AI Overviews on clearly commercial queries, but it isn't eligible for contextual matching on exploratory questions. Negative keywords and brand exclusions still apply.