If you've ever felt like Google's Performance Max gives you a new dial one quarter and quietly hides it the next, you're not imagining it. The campaign type that launched as a near-total black box has spent 2025 and 2026 handing advertisers real visibility and a few genuine controls. The catch is that "more transparency" and "more control" are not the same thing, and Performance Max keeps blurring the line.

Here's our honest read on what actually changed, what's worth your attention, and what we still would not trust Performance Max to do on a local-services account.

Campaign-level negative keywords finally arrived (with a big asterisk)

This is the headline change. In January 2025, Google announced that campaign-level negative keywords were rolling out to all Performance Max advertisers, and the cap was later raised to 10,000 per campaign to match Search campaigns. For years you couldn't tell PMax not to show on a term. Now you can.

But read the fine print, because it matters enormously for a plumber or a roofer. Multiple independent analyses found that those negatives only apply to Search and Shopping inventory — they do nothing on YouTube, Display, Gmail, Discover, or Maps.

The negatives you add only police part of your spend. Industry estimates put 40–70% of Performance Max budget on non-search channels that campaign negatives can't touch.

So yes, add your negatives. But understand you're fencing off the front yard while the back gate stays open. For the rest, you're stuck with placement and content exclusions, which are blunter instruments.

Channel-level reporting: you can finally see, but not steer

The other big 2025 win was channel-level reporting. Google announced it on April 30, 2025, detailed it at Google Marketing Live that May, and rolled it out broadly later in the year. You can now break a PMax campaign down across Search, YouTube, Discover, Gmail, Display, Search partners, and Maps — clicks, conversions, and cost per channel.

That's a real improvement. We used to guess where PMax was spending; now we can look. But notice what Google's own announcement says: the system "focuses on your main conversion goals" and optimizes across channels automatically.

Reporting is not the same as a lever. You can now watch YouTube eat a third of your budget — you just can't tell Performance Max to stop.

For a local-services advertiser whose conversions are phone calls and form fills, seeing that a chunk of spend went to brand-unsafe Display placements is useful for the next strategy conversation. It is not a control you can pull this afternoon.

The smaller wins worth knowing

A few other 2025 additions are genuinely helpful:

  • Search terms reporting came to Performance Max in 2025, so you can finally see the actual queries triggering ads — the single biggest step toward de-blackboxing the campaign type.
  • Brand exclusions got more granular, though mainly for retail advertisers with product feeds: per Google's January 2025 post, you can now exclude brand traffic from Search text ads while keeping it on Shopping.
  • In November 2025, Google added Waze ads to the PMax inventory mix — more reach, and one more place your budget can quietly go.

What Google says is coming in 2026

Per Google's own 2026 update post, the roadmap leans toward transparency again: first-party audience exclusions (so you can stop spending to re-acquire existing customers), budget reporting that projects end-of-month spend, fuller audience reporting by age and gender, and network segmentation inside placement reports for brand-safety review.

We'd flag these as promising but unproven. Google lists no firm rollout dates, and the first-party exclusion is the only item on that list that's an actual control rather than a clearer window.

What we still wouldn't trust it to do

For a $1M–$20M local-services business spending real money each month, the honest assessment is this. Performance Max in 2026 is far more legible than it was — and still not steerable enough to run unsupervised.

It still won't let you apply keyword-level control across the channels where most of the budget lands. It still optimizes to the conversion goal you feed it, which means a poorly defined goal — "form submissions" instead of "booked jobs" — quietly trains it to chase the wrong outcome. And it still decides the channel mix for you.

The fix isn't a setting. It's the goal you feed it. If your conversion signal stops at the lead, PMax optimizes to leads — not to revenue.

That's the part the new dials don't solve. The transparency features tell you where the money went; they don't tell the machine what a good job is worth. Closing that loop — from ad spend to booked jobs to revenue — is still on you and your data, not on Google's roadmap.

If you're running Performance Max and want a straight read on whether it's spending toward booked jobs or just leads, we'll take a look. Request a consultation →