Exact Match Doesn't Work in AI Overviews: How Match Types Are Changing in 2026

If you're running a Google Ads account built on tight exact match keyword lists, you've got a blind spot the size of half the search results page. AI Overviews now appear on 48% of all Google search queries - a 58% increase year over year. And here's the part most PPC managers still haven't internalized: while both exact and broad match keywords can be eligible to trigger ads above or below an AI Overview, only broad match keywords (or keywordless targeting) are eligible to trigger ads within AI Overviews.

That distinction is not a minor footnote. It's a structural shift in how Google determines who earns visibility in the fastest-growing section of search results. Seer Interactive's study found organic CTR dropped 61% and paid CTR dropped 68% on AI Overview queries. The traditional SERP is shrinking, and the new high-value real estate sits inside AI-generated summaries where exact match keywords simply cannot go. This post breaks down exactly what changed, why it changed, and what you need to do about it - without asking you to throw away everything you've built.

What Google Actually Confirmed About Exact Match and AI Overviews

In December 2025, Google Ads Liaison Ginny Marvin clarified a policy that had been confusing advertisers for months. Google confirmed that it cannot show ads in AI Overviews with the exact match type - a change from back in May, where Google had said otherwise.

The original May 2025 statement from Marvin had acknowledged that an ad could trigger to show "either above/below AIO or within AIO, but not both at this time," and "only the broad match would be eligible to trigger an ad to show within AIO." But by December, the picture became sharper. Marvin stated: "The presence of the same keyword in exact match will not prevent the broad match keyword from triggering an ad in an AI Overview, since the exact match keyword is not eligible to show Ads in AI Overviews and hence not competing with the broad match keyword."

Two things matter here. First, exact match and phrase match are locked out of in-overview placements entirely. Second, having an exact match version of a keyword no longer blocks the broad match version from appearing inside AI Overviews - a problem that had previously frustrated advertisers running mixed match-type setups.

Google is reinforcing a clear separation between traditional keyword matching and AI-powered intent matching. The system that determines which ads appear within AI Overviews doesn't evaluate keyword match type at all. Ads in AI Overviews are matched to Google's understanding of user intent based on not just the user query but also the content of the AI Overview - which is why Google requires AI-powered targeting solutions like broad match or keywordless targeting to match relevant ads.

Why Exact Match Was Never Built for Conversational Search

The logic behind this restriction isn't arbitrary. It's architectural.

AI Overviews are triggered by long-tail, conversational queries - things like "what's the best running shoe for high arches if I have knee pain?" You will never predict every variation of this in an exact match list. These queries don't look like traditional keywords. They look like questions you'd ask a friend who happens to be an expert.

The reality is that match types are now more of a thematic suggestion for Google rather than strict rules about specific keywords. Exact match, despite its name, hasn't been truly "exact" since 2018. Google quietly tweaked match types around 2021, so all match types - including exact match - have become more flexible. Instead of the user's query needing to exactly match your keyword, it now just needs to match the intent or meaning.

But even with close variants and same-intent matching built in, exact match still operates within a narrow interpretive window. Exact match shows your ad only for searches that precisely match your keyword or close variations of it - including misspellings, singular/plural forms, function word removal, word reordering, accents, abbreviations, and stems. That's a long list, but it still can't stretch to match the kind of exploratory, multi-intent queries that trigger AI Overviews.

The Intent Gap in Practice

Consider a query like "why is my pool green and how do I clean it." Google's own documentation shows that while this user query may not be directly commercial, its understanding of the query combined with the AI Overview content helps it detect commercial intent - and serve relevant ads for "pool vacuum cleaners" to help the user take the next step.

No exact match keyword list would include that query. Phrase match wouldn't catch it either. Only broad match - processing landing page content, other keywords in the ad group, user history, and location signals - can bridge the gap between "pool vacuum" and "why is my pool green."

Broad match is the only match type that uses all of the signals available - landing pages, other keywords in the ad group, previous searches, user location, and more - to understand the intent of both the user's search and your keyword. That signal depth is precisely what AI Overview placements require.

The Scale of What You're Missing: AI Overviews by the Numbers

The business case for caring about AI Overview ad placement comes down to pure arithmetic.

AI Overviews cover 48% of queries and reduce organic CTR by 15–61%.

Ads now appear in 25.5% of AI-generated results, up 394% year-over-year. Ads show at the bottom of approximately 25.5% of AI Overview SERPs, and Google's AI Mode infrastructure includes a ready-to-deploy ad placement called "AI Mode Bottom Ads" with 816 active experiment IDs testing monetization approaches.

Meanwhile, generative summaries fundamentally change the math of a successful campaign. When an AI Overview pushes paid ads below the fold, it triggers a chain reaction: lower CTR means fewer clicks, fewer clicks means fewer conversions.

Here's what makes the data alarming for exact-match-only accounts. Brands cited in AI Overviews earned 35% more organic and 91% more paid clicks than those not cited. Being visible within that AI-generated context drives measurably higher engagement. And if your match type strategy makes you ineligible for those placements, you're invisible during the highest-leverage moments.

Broad match is the only match type eligible for AI Overview and AI Mode placements. Phrase and exact keywords will not qualify for these new surfaces. For advertisers who spend six figures monthly on Search, the gap between "eligible" and "ineligible" for half the SERP represents real lost revenue.

How AI Max for Search Changes the Equation

Google's response to this shift is AI Max for Search campaigns, launched in May 2025 and pushed aggressively since. Unlike Performance Max, it's not a standalone campaign type. AI Max for Search campaigns is a suite of AI-powered features - not a new campaign type, but an optimization layer you can enable within an existing Search campaign.

The pitch from Google is strong. Google's internal data shows that advertisers activating AI Max typically see 14% more conversions at a similar CPA/ROAS. For campaigns still mostly using exact and phrase keywords, the typical uplift is even higher at 27%.

But independent testing paints a more nuanced picture. While Google promises 14% average conversion improvements and up to 27% lift for exact match-heavy campaigns, independent testing reveals a far more complex reality where 84% of advertisers report neutral or negative results. The SMEC study of 250+ campaigns found a 13% conversion value lift but higher CPA and unpredictable ROAS results.

Where AI Max Underperforms

Up to 63% of the time, AI Max's keywordless expansion was recycling existing coverage rather than discovering genuinely new queries. That means the "incremental reach" Google advertises is often just repackaged traffic you were already capturing through your keyword lists.

AI Max delivered a 35% lower ROAS than other match types within the same campaigns, paired with a lower average order value on the conversions it did generate. More conversions at worse efficiency isn't growth - it's inflation.

AI Max is a campaign-level toggle that overrides your match type settings. For advertisers who have spent years refining exact and phrase match keyword lists with carefully tiered bidding, that override can be unsettling. Having to set up guardrails to prevent a feature from doing damage is a sign that the feature isn't ready.

Where AI Max Delivers

For the right accounts, the results are legitimate. L'Oréal saw a 2X higher conversion rate at a 31% lower cost-per-conversion, specifically unlocking more conversions from net-new search queries like "what is the best cream for facial dark spots?"

MyConnect drove 16% more leads at a 13% lower CPA, with a 30% increase in conversions specifically from net-new queries.

The pattern is consistent: AI Max works best when there's a large pool of untapped conversational queries that traditional keyword lists can't capture. If your category involves complex purchase decisions, comparison shopping, or problem-solving queries, the opportunity is real.

The Practitioner's Match Type Framework for 2026

The old playbook - exact match for proven keywords, phrase for discovery, broad as an experiment - isn't wrong. It's incomplete. Here's how to update it without burning your account down.

Keep Exact Match as Your Profit Anchor

In 2026, exact match still gives the tightest control, even with all the AI on top of it. Google now allows close variants, but we still see better control and higher intent than any other match type.

Exact and phrase match still work for brand defense and high-visibility placements above the AI summaries, but they won't get you into the conversational layer where exploration happens. Use exact match for three specific jobs: high-intent bottom-of-funnel queries, proven converters promoted from broader match types, and brand protection.

Run Broad Match as Your AI Overviews Pipeline

Create a dedicated broad-match campaign or ad group for testing AI Overview eligibility. Establish clear budget rules so tests run long enough to reach statistical significance. Set up negative keyword lists from the start and update them weekly during the test. Use Smart Bidding with conservative targets initially to limit wasted spend.

This isn't about replacing your exact match campaigns. It's about running a parallel system that captures the 48% of queries where AI Overviews appear and your exact match keywords can't follow.

It's critical to use Smart Bidding with broad match. Every search query is different, and bids for each query should reflect the unique contextual signals present at auction-time. Smart Bidding uses these signals to ensure that you're only competing in the right auctions, at the right bid, for the right user.

Audit Your Search Term Reports Weekly, Not Monthly

Phrase match has become broader than ever. If you set up campaigns expecting tight phrase match behavior from years past, you'll find searches you didn't intend appearing in your ad group. That's why reviewing your Search Terms Report weekly is non-negotiable for the first 60–90 days of any campaign.

When you find high-converting queries from broad match campaigns, clone them into exact match and give them their own bid rules once a search drives at least 10 to 20 conversions with strong ROAS. This creates a flywheel: broad match discovers, exact match locks in the win.

Approach AI Max with Controlled Experiments

Don't flip AI Max on across your entire account. Use the experiment approach to test AI Max without risking existing performance - run for 4+ weeks to allow proper learning.

Achieving a 10% improvement with AI Max beats 84% of advertisers using it. Don't benchmark against Google's cherry-picked 263% case study.

Check that your data layer is solid - GTM server-side, offline conversion import, and CRM matching decide if Smart Bidding even works. Garbage in, garbage out.

The Reporting Problem Nobody's Solving

One of the most frustrating dimensions of this shift is measurement. Google Ads currently doesn't offer segmented reporting when ads show within Search AI Overviews.

Ads in AI Overviews are reported as "Top Ads," but you cannot isolate their performance metrics.

This means you can't definitively tell which conversions came from within an AI Overview versus a traditional top-of-page placement. As of early 2026, Google still does not provide segmented reporting for ads shown within AI Overviews versus traditional placements.

The workaround is indirect but effective. Cross-reference your broad match campaign performance with third-party SERP feature tracking tools like Adthena, Semrush, or Ahrefs that monitor AI Overview presence for your target queries. Compare conversion rates on queries you know trigger AI Overviews against those that don't. The gap tells you how much lift - or drag - the new placement is creating.

Advertisers cannot target ads specifically to AI Overview placements, nor can they opt out. That lack of control makes measurement discipline even more important.

What the Experts Are Saying About Where This Goes Next

The consensus among practitioners is directional, not absolute. Search themes and brand inclusions/exclusions are a harbinger of what's to come, while match types and negative keywords are on the way out.

Here's the uncomfortable truth about match types in 2026: Google is actively trying to replace keyword control with AI matching. AI Max is the latest push. It won't be the last. The direction is clear.

Google wants you to hand over your keywords, ads, and landing pages and let their system figure out the matching. They've been moving this way for years: close variants expanded, phrase match got broader, broad match improved with Smart Bidding.

Some experts push back against full capitulation. Sarah Vlietstra wrote on Search Engine Land that phrase match has recently fallen out of favor, and because Smart Bidding and broad match rely on multiple intent signals, these signals now match user intent more accurately than phrase match did under the same strategy. Others, like Dennis Moons at Store Growers, argue for a more measured approach: meet Google halfway, not all the way. Broad match with Smart Bidding already gives Google room to find relevant queries beyond your exact keywords.

In Google Ads, the auction isn't triggered by a keyword anymore - it's triggered by inferred intent. If you're still structuring campaigns around exact and phrase match exclusively, you're planning for a system that no longer exists.

Building an Account That Works on Both Sides of the SERP

The structural reality of 2026 PPC is a split world. Half the SERP still functions like traditional search - ten blue links, ad positions above and below, keyword-driven auctions. The other half is AI-generated summaries where intent matching, broad signals, and landing page quality determine who shows up. Your account needs to operate in both environments simultaneously. Exact match isn't dead. It's the most efficient match type for locking in proven, high-converting traffic on traditional placements. Treat exact match like a profit anchor - it locks in winning traffic before you scale wider.

But if every dollar in your account runs through exact and phrase match keywords, you've built a machine optimized for a SERP that covers a shrinking share of user attention. If your account is built only for narrow keyword capture, AI Overview-era Search may expose that limitation.

The practitioners who are winning right now run tiered structures. Exact match campaigns protect margin on proven queries. Broad match campaigns - paired with Smart Bidding and aggressive negative keyword management - serve as the discovery engine that feeds exact match. And for accounts with sufficient conversion volume and clean data pipelines, AI Max experiments probe the edges of keywordless search, capturing long-tail conversational queries that no keyword list could ever anticipate. That isn't a radical overhaul. It's an evolution. And the accounts that evolve fastest will own the ad placements that their competitors can't even access.