Is Performance Max wasting my budget?

Blended ROAS is the number that hides the answer

Most founders judge Performance Max on the account’s blended return. That number averages two different jobs into one grade.

Here’s a real example from my own book. Across two home-brand Shopify accounts over ninety days to late June 2026, the blended return came out to 21x. Nobody looking at that number asks whether anything is being wasted.

Split it by campaign type and the story changes. Standard Shopping took 18 percent of the spend and produced 52 percent of the revenue at 61x. Performance Max took 82 percent of the spend and produced 48 percent of the revenue at 12x. Per brand, PMax returned 17x on one and 8x on the other. Shopping was mostly catching buyers already close to a purchase. PMax was doing the more expensive work of finding new ones. The blended 21x hid a 5x gap between the two jobs. The full breakdown is in what a home brand’s 21x ROAS hides.

Neither number means PMax was failing. Break-even ROAS is one divided by contribution margin, so 12x clears break-even at any contribution margin above about 8.3 percent. The point is that you can’t know until you separate the jobs.

Three signals the waste is real

The conversion signal feeding it is wrong. Performance Max is an amplifier. It optimizes toward whatever conversion data you give it, so broken purchase values, duplicate events, or cart events counted as conversions send it shopping for the wrong customers at full price. When I audited Emma Grace Home, its Google Ads was optimizing toward add-to-carts instead of buyers. Broken conversion tracking is the most common finding in the audits I ran in 2026, and it’s invisible from inside the campaign screen. On two of three ecommerce accounts in my tracking-accuracy benchmark, add-to-cart and begin-checkout events made up 19.8% and 25.8% of the reported conversion value over 90 days. Nothing in the interface labels it. The column still reads “Conv. value.” I wrote up the checks in the conversion tracking library.

It takes credit for demand you already owned. Out of the box, PMax can serve on searches for your own brand name and on your remarketing pools. Those buyers were already looking for you, so the conversions are cheap and the reported ROAS climbs without PMax creating anything. My default on every account is that no campaign bids on the client’s own brand. Add the brand name and its misspellings as phrase-match negatives at the account level, which Performance Max respects, or apply a brand exclusion list on the PMax campaign. The one exception I make is when a competitor’s ad holds your brand search above your organic result, and then I move the brand negatives from the account level to a shared list applied to every non-brand campaign. The mechanics of PMax and Shopping trading credit are in why Performance Max gets credit for Shopping conversions.

One campaign carries every product and audience. A single PMax campaign with one asset group across a whole catalog gives the algorithm nothing to differentiate. Margins, price points, and intent levels all get averaged, and the budget flows to whatever converts easiest rather than what earns most. I run a six-campaign PMax matrix on catalog accounts for this reason.

How to run the split on your own account

1. Segment the last ninety days by campaign type. Note spend share and revenue share for Performance Max, Standard Shopping, and Search separately.
2. Pull the PMax search terms and search categories insight. If your own brand name or its misspellings show up with real spend, the brand negatives or brand exclusion aren’t in place. Fix that before you judge the campaign, because those conversions would have happened anyway.
3. Open the conversion-action list and read the category of every action in the bidding metric. If add-to-cart or begin-checkout is counted next to purchase, the ROAS you’re judging includes cart value.
4. Compare each campaign type against its own job. Harvest campaigns should run several multiples above your blended target. Prospecting should clear your break-even ROAS with room to spare.
5. Check the learning phase before judging anything. A new or heavily edited PMax campaign spends one to two weeks recalibrating, and cost per acquisition can run well above target during that window.

What a healthy PMax returns

In the two accounts above, PMax ran between 8x and 17x depending on the brand, while the Shopping layer ran far higher. If your PMax sits below break-even ROAS for thirty days or more with clean tracking, brand excluded, and a stable campaign, that’s waste, and the fix is structural rather than a bid change.

Even a PMax that clears break-even on paper is reporting attributed revenue, which isn’t the same as revenue it caused. On one home-furnishings brand, a 21-day geo holdout of all paid media showed 38% real lift against the 92% of revenue the attribution report called paid-influenced. That’s one brand, and the test is in the 21-day geo-holdout write-up. Run one before you scale PMax on its reported ROAS.

The fix order

Tracking first, structure second, creative third. Verify the purchase event, its values, and deduplication before touching the campaign, and take cart events out of the bidding metric. Then keep PMax off your brand and separate harvest from prospecting so each job gets judged and budgeted on its own numbers. Only then work the asset groups. Fixing them in the reverse order optimizes a campaign against data you can’t trust.

When to bring me in

Bring me in when cost per lead keeps rising and the fixes above haven’t moved it, when you can’t tell which campaigns produce booked work, or when the search terms report shows money going to searches you’d never pick.

I'm Conner Crowe, and I write every audit myself. Send the free audit request and add me as a read-only user, and within 24 hours of access you'll have a written read on where the spend leaks and what to fix first. No call needed, and nothing in your account or store changes.

Request a free Google Ads audit →

See how I manage Google Ads accounts.