Google this month launched a small experiment that adds a "Strongest match" or "Strong match" label to certain Search ads, using existing ad quality and relevance signals to badge ads it considers highly relevant - a test that is initially rolling out to a small percentage of users in the United States.

What the experiment actually is

The announcement came from Ginny Marvin, Google's Ads Product Liaison, in a post on X published on June 23, 2026, at 11:49 PM. The post had accumulated 4,820 views within hours of going live. Marvin described the experiment as something that is "starting this week," making today - June 24, 2026 - the first full day of the rollout.

The goal, according to Google, is simple: to help people instantly identify the most relevant information for their query, and to help advertisers connect more effectively with high-intent audiences. The labels - "Strongest match" and "Strong match" - are not new categories of inventory. They are badges applied to existing Search ads that already meet a certain threshold of relevance under Google's existing systems.

Two distinct label tiers appear in the experiment. "Strongest match" sits at the top tier, indicating Google's highest confidence that an ad is highly relevant to a given query. "Strong match" is a secondary designation, suggesting elevated but not maximum relevance. Neither label appears to affect ad eligibility, placement logic, or cost - at least not in the stated mechanics of this experiment.

The geographic scope is narrow. According to Marvin's post, the experiment is limited to the United States and is rolling out to only a small percentage of users. That framing - a small percentage of US users - is consistent with how Google typically scopes early-stage experiments before it decides whether to expand, modify, or drop a feature.

How the underlying signals work

The labels do not rely on new data. According to Google, the experiment uses "existing ad quality and relevance signals" to determine which ads receive a badge. Those signals map directly to the infrastructure that has governed Search ad quality for years.

Ad quality in Google's framework is an estimate of the experience users have when they see a Search ad and when they arrive at the landing page behind it. According to Google's Help Center documentation, this assessment is summarized in a Quality Score, which advertisers can monitor in their Google Ads accounts. Quality Score is presented on a scale of 1 to 10, calculated relative to the performance of other advertisers competing for similar queries.

Three components feed into Quality Score: expected clickthrough rate, ad relevance, and landing page experience. Expected clickthrough rate measures how likely a user is to click an ad when it appears for a given search term, calibrated against historical data. Ad relevance assesses how closely the ad matches the intent of the search. Landing page experience looks at how useful and relevant the page is to users who arrive after clicking.

Importantly, according to Google's Help Center, Quality Score is described as a "diagnostic tool" rather than an auction input. The documentation is explicit: "these scores are not inputs in the ad auction." Advertisers are instructed to use Quality Score to diagnose low-quality ads rather than to optimize the 1-to-10 score directly.

Ad Rank is the metric that actually determines whether and where an ad appears. It is recalculated at every auction and takes quality signals into account, but it also incorporates bid amounts and other contextual factors. The Help Center uses an analogy: "if your car engine has high efficiency, you need only little fuel to take you far. This is the same as Ad Rank, where good ad quality will increase the likelihood of a higher ad rank."

The "Strongest match" label appears to draw on this same quality and relevance infrastructure to surface - for users - which ads are performing strongly on the relevance dimension at the time of a given search. The note in Google's official documentation that "ad quality varies with the context of every search" is directly relevant here: the label is context-dependent rather than a permanent badge attached to an account or campaign.

What does not affect ad quality

Several factors are commonly misunderstood as influencing ad quality, and Google's documentation addresses these directly. Bidding - the amount an advertiser bids - may affect Ad Rank, but does not affect the assessment of ad quality itself. Account structure - moving an ad group with the same ads and keywords to another campaign or account - also has no direct impact on quality. One exception applies: moving a keyword to a new ad group that has different ad text can change quality scores because it affects user experience.

Frequency of ads and reported conversions are also listed as non-factors. How often ads show is determined by bids, budget, and keyword competition. Conversion data reported by advertisers is not fed back into the ad quality calculation. These clarifications matter for interpreting what the "Strongest match" label actually reflects: it is a snapshot of relevance at query time, not a measure of bid levels or conversion history.

Community reaction

Advertiser reaction on X, visible in the replies to Marvin's post, was mixed. Several practitioners questioned the basis of the label given broader frustration with keyword matching transparency. One commenter, identified as Blastoff Ads, argued that keyword matching "has been purposefully destroyed" and questioned how Google could claim strong matching when precision controls have been reduced over time. Another, Anthony Higman of ADSQUIRE, raised a specific operational question: "how do you actually work on improving quality score in any capacity when there are 15 headlines and four descriptions?"

That question reflects a real structural tension in the current Search advertising environment. Responsive Search Ads - now the only Search text ad format - require advertisers to provide up to 15 headlines and 4 descriptions, with the system assembling combinations dynamically. The relationship between individual asset quality and the aggregate relevance signal is not fully transparent. As PPC Land covered in February 2026, internal Google product managers have acknowledged this complexity, with Abby Butler noting that the system is "optimized to drive a lot of high quality clicks" and that predicted performance is "based on what our system knows about this format in this query, in this specific context."

The reply thread also contained a direct follow-up from Marvin herself, pointing users to Google's ad quality Help Center article and to a LinkedIn piece for additional context on how these experiments and evolving Search ad experiences work behind the scenes.

Where this fits in a year of Search ad labeling changes

The "Strongest match" experiment is the latest in a series of labeling and presentation changes Google has made to Search ad results over the past year. The sequence reveals a consistent pattern of testing how additional signals or designations change user behavior and engagement.

In October 2025, Google introduced a sticky "Sponsored results" label that remains visible as users scroll, replacing per-ad "Sponsored" designations with a single grouped header above multiple ads. That change also introduced a "Hide sponsored results" control, allowing users to collapse all text ads in a section. The company cited testing showing the new design "helps people navigate the top of the page more easily." The maximum number of text ads in a grouping remained capped at four.

September 2025 saw a separate consolidation of individual sponsored labels into a single grouped "Sponsored results" header, first documented by practitioners on September 24. In July 2025, Google removed the prominent "Sponsored" label from its "Find Related Products and Services" search ad units, replacing it with smaller disclaimer text at the bottom of the section.

In late April 2026, Google was also spotted testing an AI label on mobile search ads, with two variations observed - one with a rounded background and one with an information icon. That label provided no additional information when tapped, and Google offered no public explanation of what it was meant to signal. It remains unclear whether that experiment and the "Strongest match" label experiment are running simultaneously or whether one succeeded the other.

At a broader level, Google's May 2026 announcement of new Search ad formats for AI Mode introduced Conversational Discovery ads and Highlighted Answers, both of which require ads to be "highly relevant to the query and high quality" to be eligible. The eligibility framing in those formats - dual conditions of relevance and quality - is closely aligned with the logic behind the "Strongest match" label, suggesting a coherent underlying approach even if the surfaces differ.

Why relevance labeling is technically complex in 2026

The timing of this experiment intersects with ongoing changes to how relevance is determined in Google's ad auctions. As PPC Land reported in May 2026, Google's search terms report now shows AI-interpreted intent in some cases rather than the literal text users typed. For searches originating from AI-mediated surfaces - including Google Lens, AI Mode, AI Overviews, and autocomplete - the search term shown in reporting represents, in Google's own language, "the best approximation of the user's intent."

This has a direct implication for what "relevance" means in this context. If an ad is labeled "Strongest match" for a query where the underlying user action was an autocomplete selection, a voice search, or an AI Overview interaction rather than a direct text entry, the definition of "match" is more complex than it appears. The auction system in those cases uses AI-based ad group prioritization rather than literal keyword matching, selecting ad groups based on inferred overall intent.

The clarifications from Ginny Marvin published in December 2024 about AI Max matching added further context to this complexity. Marvin explained at that point that "we're increasingly determining relevance by inferred intent (like with Lens or AI Overviews) versus just the raw text query." Whether the "Strongest match" experiment's relevance assessment tracks literal keyword matching, inferred intent, or a hybrid of the two is not clarified in the available announcement.

Scale and advertiser implications

For advertisers who encounter the label during the experiment period, it raises a practical question: does a "Strongest match" designation influence user behavior? Click-through rates on labeled ads would be a natural metric to watch, though Google has not indicated it will publish data from this experiment. If higher-relevance-labeled ads attract disproportionate clicks, that could affect auction dynamics by concentrating traffic toward already-high-performing advertisers.

The experiment also intersects with Google's limited ad serving policy expansion, announced on June 12, 2026, which begins restricting impression volumes for advertisers who generate persistent user complaints on Google Search. That policy, which is gradually being enforced and will be fully in place by 2028, ties ad serving levels to user experience signals - the same category of signals that underlie ad quality assessments. An advertiser whose ads generate negative experience signals may face impression limits, and the "Strongest match" designation - which signals high relevance - would presumably appear less frequently for ads from such advertisers.

Higher ad quality has material cost implications. According to Google's Help Center, higher quality ads typically cost less per click than lower quality ads. For ads with low quality, the actual cost per click tends to approach the maximum CPC even when competition for the triggering search terms is low. The "Strongest match" designation, by making high-relevance ads visible to users, could accelerate the concentration of clicks toward better-optimized campaigns, further raising the performance floor that average advertisers need to clear.

What Google has shared and what remains unknown

The announcement provides a purpose and a mechanism but leaves a number of specifics open. The percentage of US users in the experiment is not disclosed. The threshold - whether relevance score, Quality Score value, or another internal metric - that qualifies an ad for "Strongest match" versus "Strong match" versus no label is not specified. Whether both label variants appear in the same experiment or in separate A/B branches is unclear. The duration of the experiment is not stated. And there is no commitment to expansion, modification, or a public outcome disclosure.

Marvin acknowledged this directly in her post: "we're always tweaking and testing things, so while I can't promise a preview for every new UI experiment, I thought it'd be fun to share the thinking behind this one before you start seeing it." That framing positions the disclosure as a transparency gesture rather than a product announcement. It is notable because Google typically does not publicize early-stage UI experiments before they become visible in the wild. The decision to share advance context suggests that this particular test was expected to surface quickly in the monitoring feeds that practitioners and SEO and SERP researchers use to track Google's search results.

Timeline

  • October 2025 - Google introduces a sticky "Sponsored results" label and a "Hide sponsored results" control globally, replacing individual per-ad sponsored designations (PPC Land)
  • Late April 2026 - Google tests an AI label attached to mobile Search ads, with two visual variants documented by Brodie Clark; no click-through information is provided when users tap the label (PPC Land)
  • May 17, 2026 - Google's search terms report documentation is updated to state that certain queries show "best approximation" of user intent rather than literal typed text, affecting how relevance is defined for AI-mediated searches (PPC Land)
  • May 20, 2026 - Google Marketing Live 2026 introduces Conversational Discovery ads and Highlighted Answers for AI Mode, both requiring high relevance and quality as eligibility conditions (PPC Land)
  • June 12, 2026 - Google expands its limited ad serving policy to Google Search, linking impression volumes to user experience signals and advertiser qualification, with full enforcement extending through 2028 (PPC Land)
  • June 23, 2026 - Ginny Marvin, Google Ads Product Liaison, announces the "Strongest match" and "Strong match" Search ad label experiment on X at 11:49 PM, describing it as starting "this week" and limited to a small percentage of US users
  • June 24, 2026 - First full day of the "Strongest match" label experiment rolling out in the United States

Summary

Who: Google, announced by Ginny Marvin, Ads Product Liaison, on X. The experiment affects Search advertisers and users in the United States.

What: A small UI experiment adding "Strongest match" or "Strong match" labels to certain Search ads, using existing ad quality and relevance signals to identify which ads are highly relevant to a given query. The labels are tiered, with "Strongest match" representing Google's highest relevance designation. The mechanism draws on Quality Score components - expected clickthrough rate, ad relevance, and landing page experience - without introducing new data inputs.

When: The experiment was announced on June 23, 2026, and began rolling out today, June 24, 2026. No end date or expansion timeline has been disclosed.

Where: Limited to a small percentage of users in the United States. The experiment applies to the standard Google Search results page. There is no indication of mobile-only or desktop-only targeting.

Why: According to Google, the goal is to help users instantly identify the most relevant information for their query, and to help advertisers connect more effectively with high-intent audiences. The broader context is a sustained effort by Google to make Search ad experiences more transparent and relevant at a time when AI-mediated search surfaces - including AI Mode and AI Overviews - are reshaping how queries are processed and how ads are matched to users.