AI Mode’s follow-up ads expand existing campaign reach

Eligible Google Ads campaigns can now appear further into AI Mode conversations without advertisers selecting a new placement. Google has begun rolling out multiturn ads: advertisements shown during follow-up questions in its conversational search experience. Existing Search, Shopping, and Performance Max campaigns can gain this exposure, while advertisers’ ability to select, exclude, and assess AI Mode separately remains limited.

The expansion uses Google’s existing advertising system, so an eligible campaign may reach someone as the conversation develops beyond the initial search. Google sells this advertising inventory and benefits commercially when more campaigns can compete for it. For advertisers, the practical question is whether the expanded reach fits their placement and reporting requirements.

Follow-up ads launch in the US as AI Mode advertising expands

The United States is the launch market for multiturn ads. “To allow ads in follow-up queries in AI Mode, we are launching multiturn ads in the US,” said Shashi Thakur, Google’s vice president and general manager of Search Ads and Ads on Google Experiences.

Ads can appear several exchanges into a conversation. Search Engine Land reported seeing them while testing AI Mode, and screenshots and a video published by Schwartz showed ads beneath successive responses, including the third and fourth exchanges. A user can therefore encounter a sponsored result after continuing the conversation several times.

Alongside the US multiturn launch, Google is expanding AI Mode advertising to 10 additional markets: Australia, Canada, India, Kenya, Malaysia, New Zealand, Nigeria, Pakistan, the Philippines, and Singapore. The announcements cover different availability: the international expansion concerns AI Mode advertising generally, while the multiturn launch applies to the US. Advertisers in Malaysia, Singapore, and the other newly added markets should assess their exposure accordingly.

Across those additional markets, Google lists English as the supported language for AI Mode and AI Overviews ads, with Pakistan limited to text ads. AI Overviews are generated summaries shown in Google Search. Google also plans to introduce AI Mode advertising in Brazil, Mexico, and the United Kingdom during the fourth quarter of 2026, followed by further expansion across countries, territories, and languages.

Okoone experts
LET'S TALK!

A project in mind?
Schedule a 30-minute meeting with us.

Senior experts helping you move faster across product, engineering, cloud & AI.

Please enter a valid business email address.

Conversation context changes how ads become eligible

Within an available market, Google says ad eligibility depends on the commercial relevance of the search or conversation. Its advertising documentation describes placements within or alongside a generated response, with selection informed by the user’s query and the response’s context. Each follow-up gives the system another chance to assess relevance, although an ad appears only when the conditions for serving one are met.

Google separates this assessment into direct query matching and contextual matching. Direct query matching finds an eligible ad from a search that itself expresses commercial intent. Contextual matching considers the question alongside the generated response, allowing the system to identify commercial relevance within a broader conversation.

Access to these approaches depends on campaign and keyword settings. Google lists AI Max for Search campaigns, broad match keywords, Shopping campaigns, and Performance Max as eligible for both approaches. Broad match allows matching beyond the precise keyword wording; exact and phrase match apply tighter matching rules and have a more restricted route into AI Mode advertising.

Campaign or keyword approach Direct query matching Contextual matching
AI Max for Search campaigns Eligible Eligible
Broad match keywords Eligible Eligible
Shopping campaigns Eligible Eligible
Performance Max Eligible Eligible
Exact and phrase match keywords Eligible when the search expresses commercial intent Ineligible

A campaign’s current keyword strategy therefore affects its access to conversational opportunities. Exact and phrase match keywords can participate through a commercially relevant search, while the other listed approaches can also use response context. Teams evaluating expanded reach need to distinguish general campaign eligibility from eligibility for each matching approach.

But using response context to select ads is separate from using advertising to shape a response. Google says paid advertising does not influence generated answers, while an answer’s context can inform ad selection. The direction matters: the response helps determine advertising relevance, and ads carry the label “Sponsored” and remain visually separate from organic material.

Existing buying tools preserve campaign controls but limit placement choice

Once a campaign is eligible, Google requires neither a separate campaign nor an additional budget allocation for AI Mode advertising. Existing Search, Shopping, and Performance Max campaigns can participate through the Google Ads auction. Conversational exposure therefore extends campaigns a team already runs.

Google says AI Mode uses the same auction system as standard Search ads and supports cost-per-click bidding, which charges for clicks, and conversion-based bidding, which optimizes toward specified actions. Text and Shopping ads can appear in Google’s AI-powered search experiences, and further formats are being tested. Advertisers retain familiar buying methods as the possible destinations for their ads expand.

The existing system also governs fees and traffic protection. Google charges no extra participation fee for AI Mode advertising, while ordinary charges under the applicable bidding model still apply. Existing invalid-traffic protections remain in place, preserving the mechanisms used to address activity such as invalid clicks.

Yet participation through this system limits placement choice: advertisers cannot target AI Mode or AI Overviews exclusively or opt out of them separately while continuing to advertise through Search. A team seeking AI Mode exposure alone cannot select that destination independently. Equally, a team that wants Search advertising while excluding those AI placements cannot make that separate selection.

Google says campaign safeguards still apply where supported, including negative keywords, brand exclusions, and asset pinning. Negative keywords restrict matching, brand exclusions restrict eligibility involving specified brands, and asset pinning fixes supported advertising elements in particular positions. Familiar auctions, billing, traffic protections, and these safeguards give teams substantial control over advertising activity, but they cannot satisfy an operating requirement to approve AI Mode independently of Search.

Reporting creates a related constraint because Google combines AI Mode and AI Overviews activity with standard campaign results. Those reports include impressions, meaning ad displays, along with clicks, conversions, costs, and search terms, giving advertisers measures of campaign activity that includes the AI placements. Conversions record the actions an advertiser measures. But the documented reporting arrangement gives teams no separate AI Mode performance view for assessing its individual contribution.

A team that must isolate a placement therefore faces a different decision from one assessing its campaign as a whole. Exact and phrase match eligibility raises a separate question about access to contextual matching. Teams should decide whether to change match types, increase budgets, or leave Search by comparing their operating requirements with campaign results.

These limits also need to remain distinct from claims about financial or operational damage. Wasted spend, extra labor, hidden charges, and performance damage are separate questions from whether a placement can be selected or assessed independently. A team can find the controls incompatible with its requirements without assuming poor results are inevitable.

Generated explanations raise a separate creative-control question

Placement choice concerns where an ad appears; Google is also testing a change to the explanation shown with a sponsored listing. In tests Google announced in May, Gemini models generate text explaining why a product or service is relevant to a user’s search. This presentation experiment is separate from the multiturn rollout.

To generate those explanations, Google says Gemini can use the query, the generated response, and public information, including advertiser websites. The advertiser’s website is therefore one input among several used to produce the accompanying text. The explanation can be generated without the team preparing wording specifically for the placement.

Google says advertisers cannot manually write, edit, approve, or opt out of these explanations, which directly limits editorial control. This limit belongs to the Gemini test and needs its own campaign assessment. Its availability is separate from the multiturn rollout, so a follow-up ad should not automatically be treated as carrying a generated explanation.

Google’s engagement figures measure different parts of the experience

Differences between ad formats also matter when interpreting Google’s engagement figures. In an October announcement, Google said average AI Mode queries were “three times longer” than traditional searches, citing its US research from May 2026. Longer queries give a system designed to use conversational information more context to work with.

Query length measures how people express a request, while click-through rate measures how often a displayed listing receives a click. Google reports similar click-through rates for sponsored and organic product listings within AI Mode, based on internal US listing data collected between August 20 and September 3, 2026. Google has a commercial interest in advertiser adoption of these placements, so the comparison should remain attributed to the company.

The comparison covers product listings across AI Mode, which differs from measuring ads served during later conversational exchanges. Click-through rate also measures a different stage from conversion performance, return on ad spend, or advertiser profitability. Similar click-through rates can therefore coexist with different business outcomes, because a click’s value depends on what happens after it.

Key takeaways for leaders

  • US follow-up rollout: Google is launching multiturn ads in the US while expanding general AI Mode advertising to 10 additional markets. Advertising teams operating internationally need to distinguish general availability from access to follow-up placements.
  • Keyword strategy shapes eligibility: AI Max, broad match, Shopping, and Performance Max support matching based on conversation context; exact and phrase match require a commercially relevant search. Paid search managers evaluating broader exposure need to assess these eligibility differences before changing campaign settings.
  • Placement controls remain limited: Eligible campaigns gain AI Mode exposure through existing auctions and budgets, with no separate placement opt-out or AI Mode performance view. Marketing owners with placement approval or reporting requirements need to assess whether these limits fit their operating rules.
  • Generated explanations limit editorial control: Google’s separate Gemini test creates explanations alongside sponsored listings that advertisers cannot write, edit, approve, or opt out of. Brand and compliance owners need to assess this constraint separately from the follow-up ad rollout.
  • Engagement figures require careful interpretation: Google’s reported similarity in sponsored and organic product click-through rates covers AI Mode broadly, rather than follow-up ads specifically. Budget owners evaluating campaign changes need conversion and profitability evidence alongside engagement figures.

Alexander Procter

November 16, 2026

8 Min

Okoone experts
LET'S TALK!

A project in mind?
Schedule a 30-minute meeting with us.

Senior experts helping you move faster across product, engineering, cloud & AI.

Please enter a valid business email address.