Ads in AI Overviews: How They Work and Who Should Bid

Google places sponsored Search and Shopping ads above, below or inside its AI-generated summaries, and it draws them automatically from existing Search, Performance Max and Shopping campaigns. The advertisers who should bid are those with high-intent queries, clean product feeds and mature Smart Bidding. As of October 2026, AI Overviews sit on top of a growing share of commercial results pages. Getting in is simple, because there is no separate campaign or keyword list. Measuring AI Overviews ads and controlling where they show is much harder. This guide covers both, along with who gains most and how paid and SEO teams should split the work.

How Google Ads work in AI Overviews

Google matches ads to the AI Overview's topic and the searcher's intent, then pulls eligible ads from campaigns the advertiser already runs. When Google first confirmed the format in May 2024, it said AI Overviews would draw on ads from advertisers' existing campaigns. That cuts both ways. You cannot fully opt out, and you need no new setup to take part. Your ad copy, product feeds and campaign settings carry over unchanged.

Where the ads appear

• Above the AI Overview: standard top-of-page Search ads. This is the most visible slot.
• Inside the AI Overview: a labelled "Sponsored" section within the summary. It carries text ads, Shopping listings with images or contextual sponsored links.
• Below the AI Overview: ads that the summary pushes down the page. They get far less visibility, especially on mobile.

How eligibility is likely decided

Google has not published the full auction logic, so practitioners work from this model:

1. Google classifies the query's intent. Commercial and transactional queries are the most likely to get ads.
2. It matches campaigns whose keywords (mostly broad match), Performance Max signals or product feeds fit both the summary's topic and the query.
3. It adjusts relevance with a few personalisation signals: approximate location, device type and user interactions.
4. Normal ad rank inputs then decide who wins the slot: bid, expected CTR, ad relevance and landing-page experience.

Rollout, markets and devices

Google rolled out ads in AI Overviews in stages. Reuters reported US testing of Search and Shopping ads in May 2024. In October 2024, TechCrunch covered Google rolling out a redesigned AI Overviews experience with ads. Through 2025, availability spread from the US to India and then to more countries. By early 2026, industry coverage described the ads as broadly available on mobile, desktop and tablet. Reports still differ, though. One October 2026 source says the newest phase started with US mobile users, so check your own placement data in each market.

Ads in ChatGPT and Gemini work differently. Both are framed as conversation-based or platform-native formats, and neither has confirmed automatic eligibility from existing campaigns. Our guide to ads on Perplexity covers another answer-engine model.

Should you bid on AI Overviews ads?

You should bid on AI Overviews ads if you compete on high-intent, revenue-dense queries and already run broad match with Smart Bidding, Performance Max or a strong Shopping feed. If your demand is mostly informational, the case is weaker. Your campaigns are already eligible, so the real decision is how hard to push bids and budget on queries that trigger an overview.

When to limit exposure

• Low-margin categories, where a below-the-fold click cannot clear your target CPA.
• Accounts with too few conversions for Smart Bidding to learn. Use manual bidding with device and audience adjustments until volume builds.
• Long B2B sales cycles that have no offline conversion imports.
• Branded queries that already convert organically. Measure cannibalisation before you raise brand bids.
• Mobile-heavy informational traffic, where ads most often sit below the summary.

How to set up and test campaigns for AI Overview placements

Campaigns earn AI Overview placements when they give Google flexible, intent-rich inputs. Restrictive manual structures tend to lose these slots, because they leave Google less room to map conversational queries to your ads.

Readiness checklist

• Conversion tracking: use enhanced conversions, plus offline imports for lead gen, so tCPA and tROAS optimise for revenue, not form fills.
• Bidding: use Maximize Conversions, Target CPA or Target ROAS, and keep Maximize Clicks for traffic goals. Test AI Max for Search for better intent matching.
• Match types: run broad match on core commercial themes, and keep exact match on brand and top converters as a control.
• Negatives: add account-level lists for modifiers such as "what is", "free" and "definition". Google can still match on inferred intent.
• Assets and feeds: fill every RSA headline and description slot. Give Performance Max asset groups first-party audience signals, and give Merchant Center titles keywords, complete attributes (size, material, compatibility) and accurate prices.
• Landing pages: keep them fast, match the ad's message, and answer the question the summary answers.

A testing framework

1. Hypothesis: for example, "Moving our comparison-query campaign to broad match plus tROAS will raise Absolute Top rate on queries that trigger an AI Overview, with no fall in ROAS."
2. Budget split: for example, 80% to the existing campaign and 20% to a Google Ads experiment arm.
3. KPIs: Absolute Top impression rate, CTR, conversion rate and ROAS or CPA, segmented by device.
4. Duration: four to six weeks, after a Smart Bidding learning period of about two weeks.
5. Decision rule: scale if CPA stays within 10% of target and Absolute Top rate rises. Roll back if CPA drifts more than 20%. These thresholds are examples, so set yours from your margins.

What you can and cannot measure

You can measure impressions, clicks, CTR, conversions and ROAS for eligible campaigns, but Google Ads has no report that isolates AI Overview ad exposure.

Proxy metrics to track

• Visibility: impression share, Top vs. Absolute Top rate, and the share of ads above versus below the overview, by device.
• Performance: CTR and conversion rate on queries that trigger AI Overviews, compared with queries that do not.
• Behaviour: query length and intent trends in search terms reports, plus qualified leads and sales.
• Owned channels: branded search volume and direct traffic, as signals that people saw the summary.

To estimate incrementality, run geo holdouts through Google Ads experiments or matched-market tests, then compare total paid plus organic conversions between regions. If organic conversions fall where paid spend rises, you are buying clicks you would have earned anyway. Our guide to tracking which queries trigger AI Overviews shows how to build the query list these tests depend on.

What AI Overview ads mean for SEO and content teams

AI Overview ads make organic citations more valuable. The summary and the ads compete for the same attention, and a cited brand gets visibility without paying for each click. Paid slots will keep multiplying around the answer, so your earned presence inside it is the part you can defend.

Coordinate paid and organic

• Share one map of the queries that trigger AI Overviews, tagged as informational, commercial investigation or transactional.
• Put paid budget on transactional queries such as "cloud security platform pricing". Put content effort on queries such as "best CSPM tools for AWS", where citations shape the shortlist.
• Align ad copy and landing pages with what the summary says about your category.
• Track who is cited and who is bidding on the same queries, and defend branded results pages where rivals appear in both.

Be citation-worthy

Pages that get cited answer the question directly. They also carry clear entity signals, such as consistent product names and author expertise, and they publish original data under structured headings. Schema helps Google parse a page, but it does not guarantee a citation. Our playbook on ranking in Google AI Overviews goes deeper.

How Tellr connects paid and earned visibility in AI search

Tellr runs paid media and earned visibility as one governed program, so your paid and SEO teams plan the queries you bid on and the answers you get cited in together. It fits alongside your existing Search and Performance Max work:

• Weekly tracking of who Google and its AI Overviews cite for your category's queries.
• Comparison pages and answer-shaped articles built to be quoted, published to your CMS.
• Category ad intelligence and ready-to-run creative.
• Reddit thread monitoring and guideline-checked replies behind an approval gate.

Tellr is a managed program built for teams spending $10k+ a month, so a small team running a single Search account will get more from self-serve tools.

Action checklist for AI Overviews ads

The fastest path is to audit eligibility, tighten your inputs, test against a control and align content with paid.

1. Map the queries that trigger AI Overviews, tagged by intent and device.
2. Confirm conversion tracking before you change bidding.
3. Run the experiment arm with written decision rules.
4. Brief content teams on the commercial-investigation queries where citations matter most.

AI Overviews ads reward advertisers who already feed Google clean data and clear intent signals. They punish accounts that bid aggressively without measuring incrementality. Bid where intent is high and the margin holds, limit exposure where attribution is weak, and make sure your brand is cited in the answer as well as advertised beside it.

FAQ