PMax May Now Hand Your Brand Searches Back, Right Before Black Friday

PMax May Now Hand Your Brand Searches Back, Right Before Black Friday
Google's updated PMax help page says brand queries may go to your Search campaigns, so check your brand budget before it runs out.

Google updated its Performance Max help documentation in early November 2026 to say PMax campaigns "may also filter out navigational or brand-specific search queries" so dedicated brand Search campaigns handle them instead. Advertisers whose brand Search campaigns are limited by budget could lose brand clicks that PMax used to absorb, and PMax ROAS may drop as cheap brand conversions leave its reporting.

The change was spotted by Hana Kobzová of PPC News Feed and written up by Barry Schwartz at Search Engine Roundtable on November 3. There was no blog post and no in-product notice. Just a new paragraph on a help page most people last read when PMax launched.

I think this is one of those doc edits that matters more than it looks, mostly because of when it landed. Three weeks before Black Friday is exactly when brand search volume climbs and brand budgets start capping out by mid-afternoon.

What the help page actually says now

The page in question is How Performance Max interacts with other campaigns in your account. The new section reads, in part: "Performance Max campaigns may also filter out navigational or brand-specific search queries." Google's stated reason is that "this ensures brand-related queries are handled by campaigns offering more precise control." It then tells advertisers to "ensure your dedicated brand Search campaigns have sufficient budget and consider adding Search Themes to the Performance Max campaign."

That paragraph wasn't there in the June 2026 archived version of the same page. The rest of the prioritization logic hasn't moved. A Search keyword that exactly matches the query (or a spell-corrected version of it) still beats PMax. When nothing matches exactly, the campaign with the higher Ad Rank serves.

Notice the word "may." Google doesn't say how PMax decides a query is navigational, whether it only applies when a brand Search campaign exists, or whether it touches Shopping inventory as well as text-style placements. So treat this as a direction of travel rather than a switch that flipped on a specific date. Honestly, that vagueness is a bit frustrating, but it's pretty standard for PMax documentation.

Why your PMax ROAS might sag (and why that's partly fine)

For years the complaint about PMax ran the other way: it hoovered up brand searches and took credit for sales that were going to happen anyway. The data on that is mixed, and worth knowing before you panic about a ROAS drop.

Haus, the incrementality testing company, ran experiments on brand terms inside PMax. As PPC Land reported, including brand terms drove more incremental revenue about half the time. But in every experiment, excluding brand lowered customer acquisition cost, by 19% to 60%, roughly 40% on average. Brands with average order values above $238 tended to do better leaving brand in.

Separately, Adalysis looked at 3,300+ non-retail PMax campaigns and about 1.2 million search terms. It found 45% of PMax search terms also showed up in Search campaigns, and where they overlapped, Search had the higher conversion rate in about 84% of cases. One small detail stuck with me: lots of PMax search terms were brand misspellings like "papajohns," which is exactly the navigational traffic this new language describes.

Put those together and the likely picture looks like this. If Google really does route more brand traffic to your brand Search campaign, PMax's reported ROAS should fall, because some of its easiest conversions were brand. Your blended account numbers may barely change. On paper that looks like PMax got worse. In practice it probably just stopped flattering itself.

A PMax ROAS drop in November might be the most honest number that campaign has ever reported.

The budget trap nobody will notice until Black Friday

This is where it can actually hurt. Plenty of accounts run a brand Search campaign on a modest, fixed daily budget, because brand CPCs are cheap and nobody wants to overspend on people who'd find you anyway. When that budget runs out at 2 PM, PMax has historically picked up some of the leftover brand demand. Not elegantly, but it caught the clicks.

If PMax now steps back from those queries, that safety net seems to get thinner. It's a bit like a goalkeeper who used to cover the near post deciding it's officially the defender's job, on the same week the defender gets subbed off at halftime. Your brand campaign runs dry, PMax declines the query, and a competitor bidding on your name gets the top slot for the rest of the day. During Cyber Week, when brand search volume can spike well above baseline, that's a lot of clicks to hand away.

We've written before about how seasonality adjustments around Black Friday can cost more than they return, and about how budget-limited tROAS campaigns behaved after Google's August 17 change. This is the same family of problem. Budget caps quietly decide outcomes that bidding strategies get blamed for.

A 20-minute brand audit before Thanksgiving week

None of this needs a restructure. Personally, I'd spend about 20 minutes on four checks.

1. Search lost impression share (budget) on your brand campaign. Pull the last 30 days. If it's above 5% on any day, your brand campaign is already capping out, and PMax was probably covering some of that gap. For brand, I'd treat anything over roughly 2-3% lost to budget as a fix-now number during Q4. Brand clicks are usually the cheapest conversions you buy.

2. Hour-of-day spend on brand. Segment by hour and look for the point where spend flatlines. If that happens before 6 PM on weekdays now, it'll likely happen before noon on Black Friday.

3. Brand-term share inside PMax. Use the search terms insights report, add up conversions in categories that contain your brand name, then divide by total PMax conversions. Track this weekly through December. If the share falls sharply, Google is doing what the help page says, and you can expect PMax ROAS to follow.

4. Exact match coverage for brand misspellings. Exact match is still the strongest priority signal in the account, per Google's ad group and asset group prioritization rules. If your brand gets misspelled a lot, add the common variants as exact match keywords so the brand campaign wins them cleanly.

Google's search themes tip is a little odd

Google's own advice says to "consider adding Search Themes to the Performance Max campaign" to capture this traffic. That reads strangely next to a paragraph about PMax filtering brand queries out. According to Google's search themes documentation, themes carry the same priority as phrase and broad match keywords in Search campaigns, below an identical exact match keyword. So adding your brand as a search theme mostly matters if you don't run a brand Search campaign at all.

My read: if you have a brand Search campaign, fund it properly and leave brand out of PMax search themes. If you don't have one, this is a decent moment to build one. And if you want PMax fully off brand regardless of what Google decides, brand exclusions for Performance Max still exist and do the job explicitly instead of probabilistically.

Anyway, a small aside. Google keeps moving brand handling around through help page edits rather than announcements, and it's starting to feel like the help center is the real changelog. Somebody should probably diff those pages weekly. (Somebody does, it turns out. Thanks, Hana.)

What I expect the December reports to show

My prediction, with moderate confidence: accounts running both a budget-capped brand Search campaign and PMax will see PMax-reported ROAS fall somewhere around 10-20% between November and December, and a lot of agencies will blame the holiday auction for it. Some of that will be the auction. A fair amount, I suspect, will just be brand conversions moving to the campaign that should have had them all along.

Which isn't a disaster. It mostly means the brand budget you set in January and never touched is about to matter more than any bidding setting in the account. I'd go look at it today, while the fix is still just a number in a box.