Performance Max for DTC: the strict playbook

Why is unguarded PMax a trap for DTC?

Because a DTC brand's cheapest conversions are its own fans. PMax optimizes toward conversions wherever they are easiest, and with no exclusions the easiest are people searching your brand name and recent site visitors. The campaign then reports a flattering return built on demand you already owned. The refill-goods write-up shows the moment of truth: excluding brand dropped the reported ROAS from 2.1 to 1.8 in a week, and 1.8 was the honest baseline everything real got built on.

The trap has a second jaw: placement drift. Left uncapped, PMax migrates spend toward whatever inventory is cheap that week, which is often Display and YouTube impressions that click occasionally and convert rarely for smaller brands. Before the channel report existed, diagnosing that drift took forensic work; now it takes a report tab, which removes the last excuse for funding a black box.

What are the launch preconditions?

• Brand exclusions applied, plus a dedicated exact-match brand search campaign to own name queries at a capped budget
• A run-rate near 30 monthly conversions for the campaign to learn from; below that, feed it a higher-volume proxy or wait
• Feed hygiene done: titles, categories, availability, per the Shopping service's week-one work; PMax inherits every feed weakness Shopping has
• Consent Mode v2 verified, since modeling feeds the bidding and EEA accounts without it lost modeling in March 2025
• URL exclusions on journal, careers, and support pages that cannot convert

How do you audit PMax with the channel report?

Monthly, three questions, ten minutes. The channel performance report, rolled out to all accounts by November 2025, splits spend and outcomes across Search, Shopping, YouTube, Display, Discover, Gmail, and Maps, and its search-terms view now matches the granularity Search campaigns have always had.

1. Where did the money go? A DTC PMax should be Shopping-heavy. A quiet slide toward Display or YouTube without a matching conversion story is drift, and it gets capped.
2. What did the search terms say? Read them like any search campaign: waste themes become negatives, and brand terms appearing here mean the exclusions have a hole.
3. What does the mix weigh? Each channel carries a different emissions factor, so the same report drives the carbon column: capping the refill account's video-heavy placements cut its estimated media emissions by roughly a third while the return rose.

How do you read PMax results without fooling yourself?

Two denominators, printed monthly. First, brand-excluded return against your margin-derived break-even, with the 2025 ecommerce average of 2.87 and the 2024 median of 2.04 as the field. Second, exact-match impression share on your core search terms, the cannibalization tripwire: if it sags after PMax launches, the campaign is winning internal auctions against your own search, and the "growth" is repackaged demand. Google's newer AI Max layer for search claims a roughly 7% conversion lift at comparable CPA or ROAS; hold it to the same two denominators, because aggregate claims are not your account.

What does PMax weigh in carbon?

More than Shopping, by design. Standard Shopping tiles sit at the light end of the 1–5g CO2e per-impression range; PMax's display and video placements sit several multiples higher per impression. That is a real cost even when the return holds, and it belongs in the same monthly row as spend, per the carbon method. The channel report turned this from estimate-by-inference into arithmetic: impressions per channel times channel factor, summed. Capping PMax's heavy-format share is the single biggest emissions lever most DTC accounts have, and in the refill case it improved the return at the same time.

The quarterly verdict discipline

Every PMax campaign carries a written verdict each quarter: scale, hold, or kill, from the two denominators plus the channel audit. No zombie campaigns surviving on inertia, no scaling on a blended number. This is the same one-metric honesty the client write-ups are built on, applied inward. A campaign that cannot survive a written verdict was never performing; it was just unexamined.