Performance Max brand exclusions: block your own brand or let PMax keep those sales?

Say most of a store's sales come from people typing the brand name. PMax picks up those brand searches and reports them as its own conversions, so its ROAS looks great, even though plenty of those buyers would have found the store through organic or the brand campaign anyway.

My view is to add your brand terms as brand exclusions in PMax and let it compete for new demand only. Then compare total conversions and revenue for the whole account before and after, not just PMax ROAS. PMax numbers will look worse for a few weeks, and that's the point, because you're no longer paying twice for the same sale.

The counterargument is that some brand searchers would have converted through a generic product ad, so excluding them loses real sales. Fair, but it depends on how many of those shoppers you can actually see in the data. If you can't tell, treat the exclusion as a test rather than a loss.

Has anyone run brand exclusions in PMax and seen total conversions actually drop?

Total-account comparison is the right call. The catch is that brand exclusions only block the terms you actually list. Misspellings, brand plus a product name, and brand plus a city all slip through, so PMax keeps picking up brand sales and the "after" numbers look cleaner than they really are. Before you change anything, pull the PMax search terms and count how much brand-containing traffic is still getting in. And compare against the same weeks last year, otherwise a seasonal dip will get blamed on the exclusion.

Agreed on the leakage, but the terms that slip through matter more than the count suggests. Brand plus product or brand plus city searches are high-intent, and a dedicated brand Search campaign usually wins them cheaply. So the exclusion list needs somewhere to send that traffic. Otherwise a dip in total conversions could just be those buyers going to competitors. Do you already run a brand Search campaign next to PMax?

Brand CPCs are often low, but the brand campaign only wins those queries if its keywords match them, so "usually wins them cheaply" skips a step. Phrase match on your brand name tends to catch brand plus product or brand plus city, so you probably don't need a separate list for those. The variants that really slip through are misspellings and odd spellings, and those are cheap to add as brand keywords. If you're tempted to blame a dip on competitors, look at the brand campaign's impression share on your brand terms first.

Impression share is a decent check, but for local services I'd start with where the brand searches are landing. With plumbers and clinics, a lot of brand traffic turns into calls straight from the Business Profile listing, and those calls don't show up as Google Ads conversions unless you track them. So a drop in PMax conversions can look like lost sales when the calls just went to Maps instead. Are you tracking calls from the profile, or only form fills and site actions?

Calls are the first thing I'd fix, and I'd make it a rule for the exclusion decision: don't exclude brand terms in a market until you can see every conversion type that brand traffic produces, calls and direction requests included. If you can't, the test measures the wrong thing. The sale still happens, just somewhere your reports can't see it.

In practice, turn on Business Profile call and direction tracking and let it run for a few weeks before you change anything. Then compare total calls, form fills and site actions together, not Google Ads conversions alone. If the total stays within your normal week-to-week swing after the exclusion, the exclusion is doing its job. If it drops and the profile calls drop with it, you've learned something real, and you can decide whether those brand searchers are worth bidding on in a separate campaign.

Agreed on the rule. Before you trust any call numbers, though, check the minimum call duration on the Google Ads call conversion. Calls shorter than the threshold don't count, and brand searchers often make short calls, like checking opening hours or whether you have a product in stock before driving over. So a drop in Ads call conversions after an exclusion might just mean fewer people staying on the line long enough to count, not fewer people calling.

Pull calls and directions from Business Profile insights for the same weeks and use that as your baseline, then compare it with the Ads call conversions. If the two move together, your tracking is picking up the same behaviour. If they split apart, look at the duration threshold and whether the listing is actually linked to the account before you blame the exclusion.

Setting the duration threshold right still won't make the two numbers match. Ads call conversions only count calls that start from an ad or a call extension. If a brand searcher taps your Maps listing and rings from there, that call never shows up in the Ads number, so the two series can diverge for structural reasons, not because people hung up early. Before you read a gap as a tracking fault, check which surface each call came from. For listing-driven demand, the Business Profile count is the one to trust.

Before you read the two series side by side, write down what share of brand conversions comes from the listing versus Ads. A few weeks of data is enough for that. If the listing is more than half of brand contacts, judge the exclusion on the combined total only, because the Ads number will mislead you. When PMax loses a click, the same buyer can still call from Maps, so the Ads report shows a loss that never actually happened.