Performance Max spent years as a black box: you fed it budget, assets, and a goal, and Google decided how to split spend across Search, Display, YouTube, Gmail, Discover, and Maps without telling you the breakdown. That is changing. In January 2026 Google rolled channel-level reporting out to all customer IDs and documented it in Google Ads API v23, and it activated account-level placement exclusions that apply across PMax, Demand Gen, YouTube, and Display at once. A July 2026 alpha went further, letting some advertisers switch off search partners and display entirely. For B2B SaaS, these controls finally make PMax diagnosable — and defensible.
This matters more for SaaS than for almost any other advertiser, because SaaS conversions are scarce and costly, and the classic PMax failure mode is that budget drifts to cheap Display and YouTube inventory that produces flattering conversion counts but little real pipeline. Until now you could suspect that was happening but not prove it. This guide covers what the new reporting shows, how to read the channel split for a B2B SaaS account, how to use the account- level placement exclusions to cut waste, and how to decide — with data, not dogma — whether a channel deserves your budget at all.
What Google actually shipped
Three distinct capabilities landed, and it helps to keep them separate. First, channel-level reporting: a Performance Max campaign now reports impressions, clicks, cost, and conversions broken out by network — Google Search, Search Partners, Gmail, YouTube, Display, Discover, and Maps — rather than as one aggregate. This reached all customer IDs in January 2026 and is available in the interface and, via API v23, programmatically. Second, a "where ads showed" placement report that shows which network each placement belongs to, so you can see not just that your ad ran on a given app or site but whether it came through Display, Search Partners, or YouTube. Third, controls to act on that visibility.
The controls come in two forms. Account-level placement exclusions, activated in January 2026, let you block specific placements — individual apps, websites, YouTube channels, or content categories — across all your Performance Max, Demand Gen, YouTube, and Display campaigns simultaneously, so you maintain a single list instead of duplicating it. Then, in a limited alpha that surfaced in late July 2026, some advertisers gained the ability to switch off search partners and display for a Performance Max campaign wholesale, alongside household- income exclusions. Not every account has the alpha, but every account has channel-level reporting and account-level placement exclusions today, and those two together are enough to run PMax with real oversight.
Reading the channel split for B2B SaaS
The first thing to do with channel-level reporting is compare where the money went with where the qualified conversions came from. Pull the campaign's spend by channel next to its conversions by channel, and look for the mismatch that plagues SaaS accounts: a large share of budget on Display or YouTube, and a large share of genuine conversions on Search. A split like 60% of spend on Display against 80% of conversions on Search is the signature of a campaign optimising toward cheap inventory. The blended cost per conversion looks fine, which is exactly why the problem survived so long — the average hides the fact that the expensive, scarce Search conversions are subsidising a lot of low-value Display activity.
Read this alongside conversion quality, not just conversion count, or you will draw the wrong conclusion. If your PMax campaign optimises on raw form fills, Display and YouTube will look productive because they generate cheap leads — leads that disproportionately fail to qualify. The channel split only tells the truth once you are feeding pipeline-quality signals back into the campaign. This is the same discipline behind retraining Performance Max off micro-conversions: until the algorithm optimises toward SQLs and opportunities, the channel report shows you where cheap leads came from, not where pipeline came from. Combine the two views and the decision about each channel becomes obvious.
Using placement exclusions to cut waste
Once channel-level reporting and the placement report show you where budget is leaking, the account-level placement exclusion list is how you plug it. Review the "where ads showed" report on a regular cadence — weekly while you are cleaning up an account, monthly once it is stable — and exclude the recurring offenders: mobile-game and utility apps, made-for-advertising sites, auto-playing video placements on irrelevant content, and YouTube channels with nothing to do with your category. Because the exclusion applies across PMax, Demand Gen, YouTube, and Display at once, one well-maintained list protects your whole account instead of requiring you to replicate the work campaign by campaign.
Placement exclusions are hygiene, not strategy, so pair them with structural fixes. Excluding junk placements one at a time is a losing game if the campaign's incentives keep pushing it toward low-value inventory; the durable fix is better signals and, where warranted, walling off your highest-intent demand. For the mechanics of placement-level control across campaign types, see our guide to placement targeting for B2B SaaS. And when you audit a PMax campaign end to end — assets, signals, cannibalisation, and now channel spend — work from a repeatable process; our walkthrough of how to audit Performance Max campaigns now has channel-level data to make the audit far more precise than it used to be.
Deciding whether a channel earns its budget
With the data in hand, you can make the decision PMax previously denied you: does each channel earn its keep? The answer is not a blanket "turn off Display and YouTube" — those networks can play a legitimate upper-funnel role for SaaS, seeding awareness that pays off in later branded search. The answer is conditional on what the channel produces once you measure on pipeline. If Display and YouTube are generating SQLs or demonstrably influencing pipeline at an acceptable cost, keep them. If they only produce cheap form fills that never qualify, they are both wasting budget and poisoning your Smart Bidding signal by teaching it that low-quality leads are the goal.
For accounts with the July 2026 alpha, the channel on/off switch makes acting on that verdict trivial — turn off the network that fails to produce. For everyone else, the levers are account-level placement exclusions to constrain the worst inventory, tighter conversion goals so the algorithm chases quality, and the more fundamental option of separating your high-intent Search demand into a dedicated Search campaign where you keep full control. That last move also addresses the risk that PMax poaches conversions your brand and Search campaigns would have won anyway — a problem we cover in Performance Max brand cannibalization for B2B SaaS. Channel-level reporting is what makes all of these decisions evidence-based rather than guesswork.
Feeding better signals, not just cutting
Cutting waste is only half the job; the other half is improving what you feed the campaign so it makes better allocation choices on its own. Performance Max spends where it expects conversions, so the quality of your conversion signal and your audience inputs directly shapes the channel split you will read next month. If you tighten your conversion goal to true pipeline events and supply strong first-party audience signals, PMax has a better basis for favouring the inventory that actually converts qualified buyers — which over time shifts spend toward Search and high-value placements without you having to fight it placement by placement.
This is where reporting and inputs form a loop. Read the channel split, cut the obvious waste with exclusions, then improve the signals and audience inputs, and re-read the split a few weeks later to confirm the mix moved in the right direction. For the input side of that loop, our guide to tuning PMax search themes and audience signals covers what to feed the campaign. And if the channel report ultimately convinces you the campaign type is wrong for your motion, revisit the more basic question in should B2B SaaS run Performance Max. The new transparency does not just help you tune PMax — it gives you the evidence to decide whether to run it at all.