On September 23, 2026, Google shipped Google Ads API v25.2, and the headline change for B2B SaaS is a new competitive signal: percentile benchmarks. The BenchmarksService.GenerateBenchmarksMetrics endpoint now returns where your account ranks as a percentile tier against all advertisers in your category — not just your own historical averages — in a new CustomerMetrics.percentile_metrics field. For the first time, Google will tell you programmatically whether your CTR, CPC, or conversion rate sits in the bottom quartile or the top decile of comparable accounts.
That is a genuinely useful read, and also an easy one to misuse. A low percentile on a raw platform metric does not mean your account is failing, and a high one does not mean it is winning — B2B SaaS metrics only make sense against pipeline and CAC, not against a broad advertiser pool. This guide covers exactly what v25.2 exposes, how it differs from the benchmark columns already in the interface, how a head of growth should read a weak tier, and what to do before you touch a single bid. It is a minor, non-breaking release (per the release notes), so nothing changes automatically — the data simply becomes queryable.
What v25.2 actually exposes
The core addition is percentile tiers on the benchmarks service. Where the API previously returned category average metrics, GenerateBenchmarksMetrics in v25.2 can now return your customer’s competitive standing among other advertisers as a percentile tier, scoped with category filters, delivered in the percentile_metrics field as a BenchmarksCustomerPercentileTier value. The practical effect is that a team can pull a programmatic answer to “how do we compare to everyone else running ads in our space” and wire it into a dashboard or an alert, rather than eyeballing a single average column.
It is worth being precise about what it is not. This is not account-level change — nothing in your bidding or delivery shifts because v25.2 exists — and it is not a verdict on account health. It is a distribution-aware comparison on a handful of platform metrics. The other changes in the release are narrower: asset-group-level URL tracking in Performance Max, a Smart-to-PMax paused draft path, two new “bid too low” recommendations, and some vertical reporting fields. Of those, the PMax URL tracking is the one most B2B SaaS teams will care about, because it affects how clicks from different asset groups get tagged — relevant if you are already wrestling with Performance Max channel reporting and placement exclusions.
Why this beats the benchmark columns you already have
Google Ads has shown benchmark CTR and benchmark CPC columns in the interface for years, but they are single averages, and an average hides the shape of the distribution. A CTR that is 10% below the category average could be a hair under the median or deep in the bottom decile — the column looks identical either way. Percentile tiers add exactly the information that average was missing: which band you occupy. “Our conversion rate is below benchmark” becomes “our conversion rate is in the bottom 25% of software advertisers,” and only the second version tells you whether this is a rounding error or a real gap worth a week of work.
Because the tiers come through the API, they also escape the interface. You can fold them into the same board-level reporting where you already track the metrics that matter, rather than leaving them in a column nobody opens. If you are building that reporting layer, the percentile read slots naturally alongside the five Google Ads metrics SaaS boards actually care about — as context for those numbers, not a replacement. A percentile is a comparison; your CAC and pipeline are the outcome.
The B2B SaaS caveat: broad benchmarks lie about narrow accounts
The “all advertisers” pool is the catch. Percentile benchmarks compare you against a broad category, and B2B SaaS accounts are structurally unusual: long, multi-touch sales cycles, high-ACV deals, and conversions that are often imported from a CRM weeks after the click. Those traits distort the exact platform metrics the benchmark ranks. A B2B SaaS account that deliberately bids on expensive, high-intent terms like “[category] pricing” will post a high CPC and may land in an unflattering percentile — while generating the cheapest qualified pipeline in its niche. The benchmark cannot see the pipeline; it only sees the click price.
The same trap runs the other way on conversion rate. An account optimised toward easy top-of-funnel actions — ebook downloads, newsletter sign-ups — can rank in a strong conversion-rate percentile while producing almost no SQLs. That is the core measurement problem we keep returning to: a healthy-looking platform metric that is disconnected from revenue. It is the same reasoning behind optimising for SQLs, not leads, and it is why a percentile tier should start an investigation, never end one.
How to read a weak percentile tier
When a metric lands in a low tier, trace it to a cause before changing anything. A low CTR percentile usually points at ad relevance, match-type sprawl, or weak responsive search ads for the queries you are actually appearing on — a targeting and creative problem. A high CPC percentile (you are paying more than peers) is only a problem if pipeline is thin; if your CAC and SQL volume are healthy, you are simply competing in a premium auction and the spend is earning its keep. A low conversion-rate percentile points at landing-page fit, a conversion-tracking gap, or bidding toward the wrong action.
The decisive move in every case is to cross-reference the benchmark against your own revenue data. A below-median tier that lines up with rising CAC and falling SQLs is a real finding; a below-median tier sitting on top of healthy pipeline is noise you should ignore. This is also where Google’s own automated nudges can mislead — a percentile gap often triggers “bid too low” recommendations, and those deserve the same scrutiny as any other, which is the discipline we lay out in how to evaluate Google Ads AI recommendations. A benchmark that recommends you spend more is still a benchmark, not a business case.
Fold it into measurement, not reflexes
The right home for percentile benchmarks is your standing measurement review, not a weekly bid-tweaking ritual. Pull the tiers quarterly alongside your spend and pipeline reporting, use them to flag which metrics are genuine outliers versus which are artefacts of your bidding strategy, and feed only the real outliers into a prioritised fix list. Paired with a proper view of what you spend and what it returns — the kind of peer-set reading in our Google Ads spend benchmarks report for B2B SaaS and the ACV-segmented numbers in our vertical and ACV benchmarks — a percentile tier stops being a vanity comparison and becomes a way to spot where you are genuinely leaving money on the table.
Used that way, v25.2’s benchmarks are a small but real upgrade: for the first time Google will tell you, as a hard number, whether your account is below par for your category. The judgement it cannot make for you is whether that gap reflects a problem or a deliberate, pipeline-positive choice — and that judgement is the whole job.
Turn a low percentile into a plan
If the benchmark says your account is in a low tier and you are not certain why, that uncertainty is the finding. The accounts that benefit from a competitive read are the ones that can connect it to their own CAC, SQL, and pipeline data; the accounts that get hurt are the ones that react to a weak tier by raising bids on traffic that was never going to convert. A structured Google Ads audit cross-references signals like these against your real revenue so you learn whether a below-median benchmark is a problem to fix or a strategy to defend — and the free 10-point audit checklist is a fast way to start that comparison yourself. Either way, treat the new percentile tier as a prompt to look, not a reason to react — the same principle behind what a SaaS Google Ads audit is.