Google and ChatGPT Move Conversational Ads Toward Platform-Managed Bidding at Scale
Executive Summary
Conversational discovery just became a managed-bidding channel. Google is taking AI Mode ads to more than 10 new countries and serving them across follow-up turns, while ChatGPT Ads replaces Max Bid and Max CPC with average cost targets. Your leverage shifts from bid mechanics to conversion and margin data. Before Q1 2027 planning locks, ring-fence a test budget from search, run margin-weighted cost targets with an offline conversion feed, and checkpoint at 90 days.
The Signal
Google is extending English-language ads in AI Mode to more than 10 new countries and launching multiturn ads in the U.S., which serve ads across follow-up queries within a single AI Mode conversation rather than only at the initial prompt. Separately, ChatGPT Ads is retiring Max Bid and Max CPC controls in favor of average cost targets, with staggered cutover dates for new campaigns, existing campaigns and API partners. Meta Platforms paused its updated business and ad terms, which were due to take effect Nov. 20, leaving existing crawl rights for advertisers in place for now. All three developments were reported Nov. 10. Together they show the two dominant AI answer surfaces moving to monetize conversational discovery at scale while shifting bidding control toward platform-managed optimization.
What Changed
Advertisers can now buy placement inside multi-step AI conversations, not just the first query, and across more than 10 additional markets beyond the U.S. ChatGPT Ads shifts buying from manual bid ceilings to average cost targets, so the platform's optimization decides auction-level pricing. Conversational ad inventory is becoming a scalable, internationally available channel with algorithm-led bidding, replacing a test-budget experiment with manual controls.
Why It Matters
The sharpest implication: conversational discovery just became a managed-bidding channel, and managed bidding favors whoever feeds the machine the best inputs. Manual bid caps were the last place a media team could encode its own view of what a conversation is worth. With ChatGPT Ads moving to average cost targets, that judgment is delegated to the platform, which now sets auction-level pricing against a target you supply. Your leverage shifts from bid mechanics to conversion definitions, margin data, and creative volume. Multiturn ads change inventory economics more than they change reach. A single AI Mode conversation can now carry several ad opportunities, which means higher ad load per session and a pricing model that rewards advertisers who stay relevant as intent sharpens across follow-ups. Expect the late-conversation, high-intent turns to clear at a premium, and generic first-prompt coverage to lose its edge. What gets devalued is the test-budget posture. Treating these surfaces as a controlled experiment with hand-set ceilings stops being a viable way to learn once the platform optimizes against targets you cannot see inside. Also devalued is any plan that treats AI answers as organic-only: the paid layer is now international and conversation-deep. The why behind the why: both platforms need ad revenue to underwrite inference costs, and algorithm-led bidding is the fastest route to scale without building a manual-control ecosystem. Meta Platforms pausing its Nov. 20 terms signals the same pressure from another side. Crawl and data rights are the contested input to all of this, and the rules are not settled. The competitive move is to enter now with clean cost-target inputs before the learning curve gets expensive.
Marketing Impact
media
Bid-level control on ChatGPT Ads disappears in favor of average cost targets, so media teams stop managing auctions and start managing targets, conversion signals, and budget pacing. Multiturn ads in Google AI Mode require follow-up-intent creative and keyword themes, with performance read at the conversation level rather than the query level.
ecommerce
Multiturn ads reward feeds that stay relevant as shoppers narrow from category to specific product across follow-ups. Margin, inventory, and availability data become bidding inputs, since cost targets optimize toward whatever value signal you supply. Thin or stale feeds will lose the later, higher-intent turns.
marketing ops
Meta Platforms pausing its Nov. 20 terms leaves crawl rights intact only until it republishes them. Ops and legal need a contract-change watchlist, a documented crawl-access position, and a fast route to review revised terms. Separately, cutover dates for ChatGPT Ads campaigns and API partners demand a migration calendar this quarter.
martech
Teams running ChatGPT Ads through API partners face their own cutover date, so bid-management and reporting integrations built around Max CPC logic need auditing now. Conversion APIs, offline conversion uploads, and value-based signals become the primary tuning mechanism and need to be clean before Q1 2027 budgets load.
The Exploit
Opportunity
Average cost targets reward whoever feeds the optimizer the best signal. Performance leads can capture this by defining conversions at the margin level rather than revenue, uploading offline and LTV-weighted outcomes, and building 15-20 creative variants matched to early versus late conversation turns. Multiturn inventory means late-turn, high-intent placements can be won before ad load repricing.
Risk
Delegated bidding on opaque auctions can overspend against a poorly specified target. Attribution for conversational ads is immature, Meta Platforms' paused terms show data-rights rules remain unsettled, and staggered ChatGPT Ads cutover dates may disrupt early campaign structures.
The Move
Before Q1 2027 planning locks, the head of performance media should carve a ring-fenced test budget out of search for ChatGPT Ads and Google AI Mode, running on margin-weighted cost targets with an offline conversion feed. Checkpoint at 90 days: cost per margin-qualified conversion against paid search, before scaling.
First-Mover Advantage
Gains
Cost-target learning phases consume budget at today's low auction density. Early entrants bank trained campaigns, clean conversion histories and creative performance data across multiturn sessions, while later entrants pay higher clearing prices to teach the algorithm the same lessons.
Risks
Delegated bidding on opaque auctions can overspend against a poorly specified target. Attribution for conversational ads is immature, Meta Platforms' paused terms show data-rights rules remain unsettled, and staggered ChatGPT Ads cutover dates may disrupt early campaign structures.
Window
Roughly 6-9 months, until about mid-2027. The window closes when conversational auction density rises, late-turn inventory reprices upward, and agencies standardize cost-target playbooks that erase the learning-curve edge.
Winners & Losers
Winners↑
Performance teams with clean conversion and margin data
Average cost targets hand auction-level pricing to the platform, so the quality of the signal you feed it becomes the main lever on efficiency. Teams that already pipe offline conversions, lifetime value, and product-level margin into their ad accounts can set targets against real profitability while competitors optimise against platform-default proxies. Audit conversion definitions and margin feeds before ChatGPT Ads cutover dates hit your account, not after.
Creative production teams running generative pipelines at volume
Multiturn ads in Google AI Mode create several ad opportunities per conversation, and the relevant message changes as intent sharpens across follow-ups. That rewards advertisers who can field many variants mapped to conversation stages, which is a volume problem generative pipelines solve and manual production cannot. Build a library keyed to early-exploration, comparison, and decision-stage prompts so late-turn, high-intent placements are not served generic copy.
Global brands with multi-market English-language footprints
The expansion of AI Mode ads to more than 10 new English-language markets turns conversational inventory from a U.S. test into a multi-country channel without requiring localisation spend. Brands with existing English assets and regional P&Ls can stand up coverage quickly while local competitors are still deciding whether the channel is real. Prioritise markets where your search share is strong and set up market-level reporting now, because early cost-target learning accrues per campaign.
Media teams with a measurement discipline for AI-mediated funnels
When bidding control moves to the platform, advantage shifts to the team that can prove incrementality and tell the platform's optimiser what a good outcome actually is. Teams running holdouts and conversion-lift tests on conversational placements will set smarter cost targets and defend budget with evidence, while others rely on platform-reported numbers. Stand up geo or audience holdouts on AI Mode and ChatGPT Ads in Q1 2027 so the learning phase is not read at face value.
Losers↓
Media teams built around manual bid ceilings
Max Bid and Max CPC were the last place a buyer could encode their own view of what a conversation is worth. Replacing them with average cost targets moves that judgment to the platform's optimiser, which devalues bid-mechanics skill and exposes teams whose edge was hand-tuned ceilings. The defensive move is to retool toward conversion definitions, budget pacing, and creative testing, and to track how realised CPCs drift against targets once cutover happens.
Advertisers treating AI answer surfaces as small controlled experiments
A fixed test budget with hand-set caps stops producing reliable learning once the platform optimises against targets whose internals the advertiser cannot see. Meanwhile, ad load per AI Mode session is rising and late-conversation turns will likely price at a premium, so late entrants face a steeper learning curve at higher cost. Replace the pilot posture with a committed budget line and a defined learning window, or accept that competitors will have trained the algorithm first.
Brands whose AI-search plans assume organic-only visibility
Strategies that rely on earning citations in AI answers without a paid layer now face an international, conversation-deep ad product sitting on the same surfaces. Paid placements across follow-up turns will compete with organic mentions for attention, and the unresolved crawl and data rights question, underlined by Meta Platforms pausing its Nov. 20 terms, means organic access rules remain unstable. Pair GEO work with a funded paid presence in AI Mode so visibility does not depend on a single channel.
Strategic Outlook
Expect cost-target bidding to become the default on every conversational surface by mid-2027. Manual caps are the exception that vanishes once a platform has enough conversion data to optimize without them, the same path Google Ads took from manual CPC to Smart Bidding. Multiturn ads will migrate beyond the U.S. quickly, since Google is already broadening AI Mode ads internationally, and ChatGPT Ads will likely follow with comparable conversation-depth formats. Early pricing inefficiency will close fast: expect late-turn, high-intent placements to clear at a premium by the end of Q1 2027 as budgets migrate in. Meta Platforms will return with revised terms once its data-rights position is settled, and the contested ground will be crawl and training access rather than ad mechanics. Agencies and in-house teams that built conversion-signal hygiene before the learning curve got expensive will hold a measurable cost advantage through the first half of 2027.
Sources
Search Engine Journal
Google Expands Ads In AI Mode To 10+ Countries, Adds Multiturn Ads via @sejournal, @brookeosmundson
Search Engine Journal
Meta Pauses Its Updated Business And Ad Terms Due Nov. 20 via @sejournal, @MattGSouthern
Search Engine Journal
ChatGPT Ads Is Replacing Max Bid & Max CPC With Cost Targets via @sejournal, @MattGSouthern