Performance Max has offered zero channel influence since launch. You set a budget, the algorithm decides where the money goes, and you get a breakdown of where it went with no lever over where it goes next.
That's starting to shift. According to Search Engine Land, Google is testing a new "Channels" setting inside PMax that lets advertisers make positive or negative adjustments for individual channels — Search, YouTube, Display, Discover, Gmail, and Maps.
The feature is in alpha. Most accounts won't see it yet. But understanding what it actually does matters even if you don't have access, because it changes how you should be thinking about PMax structure going forward.
It's not budget allocation
The channel controls don't let you say "put most of my budget on YouTube and less on Search." That would contradict the core design of PMax, which is built on letting the algorithm find the most efficient mix across all channels simultaneously.
What the adjustments do is signal how much CPA you're willing to accept per channel. A positive adjustment on YouTube tells the algorithm it can accept a higher CPA for conversions attributed to that channel — effectively loosening the constraint. A negative adjustment tightens the acceptable CPA floor there, making PMax bid more conservatively on that surface.
The algorithm still decides what to spend. You're changing the optimization criteria, not directing the distribution.
Why that distinction matters
Most advertisers who say "I want more YouTube spend" actually mean something more specific: they're willing to accept a higher CPA on YouTube because they value the upper-funnel exposure, even when attribution is thin. The channel adjustment controls express exactly that. A positive adjustment on YouTube is a direct statement of that trade-off to the bidding system.
The negative adjustment is equally useful and less discussed. If your PMax account is attributing a high share of conversions to Display and you're skeptical of that attribution quality, a negative Display adjustment tightens the CPA floor on that surface. You're communicating skepticism to the algorithm without having to split the campaign or exclude the channel entirely.
Whether those signals actually shift channel delivery in a meaningful way is genuinely uncertain at this stage. Alpha testing exists precisely because Google doesn't fully know either.
The broader pattern worth watching
July 2026 brought a separate alpha: a "Partners" setting that lets PMax campaigns opt out of Search Partner Network and Google Display Network in a limited test. Now August adds channel-level CPA bias controls. Two consecutive months of new PMax influence mechanisms is not a coincidence.
Google is incrementally adding advertiser levers to a campaign type that launched with none of them. The direction is clear even if the pace and final shape aren't.
That said — alpha features in Google Ads have both graduated to production and disappeared quietly with no announcement. Don't restructure your PMax strategy around a control you can't access. Build your baseline.
What to do before and after you get access
If you have alpha access, test a single positive adjustment on one channel where you believe demand exists but the current CPA ceiling is limiting delivery. Start with the channel where your performance data already suggests underinvestment. Don't stack adjustments across four channels at once — one variable at a time, let it run through a full learning cycle.
If you don't have access yet, the preparation is the same work that PMax has always rewarded: complete asset groups covering all six channels, clean Merchant Center feeds, and verified conversion tracking with no pipeline gaps. Channel adjustment controls will compound the advantage of well-structured campaigns. A broken asset group baseline with a new adjustment layer on top doesn't improve things.
The most actionable task for most accounts right now is pulling the last 90 days of channel performance data in your PMax campaigns. Build a view of which channels are spending, at what CPA, and whether that distribution makes sense given where your customers actually are. When channel controls reach general availability, that baseline is what tells you whether any adjustment actually changed anything.
PMax is still a machine that allocates budget toward efficiency. These controls add a preference layer — not a hard override. Whether that preference layer gives you the control you actually need depends on how misaligned your current channel distribution is with your goals.
If your PMax channel breakdown looks wrong but you can't tell whether it's an asset gap, a bid issue, or something else, the free Gromerce audit surfaces the key structural signals in a few minutes.
Sources: Search Engine Land, PPC Land, August 2026
