PMax Brand Exclusions: Stop Hospital Brand Searches Flattering Your CPL

This post is for marketing teams who run Performance Max for a hospital or clinic group and suspect that the cost per lead looks better than the business feels. A common cause is simple: people who already know the hospital type its name into Google, click an ad, and book. That is valuable demand, but it is not the new-patient demand Performance Max is usually bought to create. When both sit in one number, the campaign gets credit for patients who would likely have found you anyway.

The aim here is to separate the two honestly. We cover what Google's brand exclusions do and do not cover, how account-level negatives differ, how to build the split, and how to read the two numbers afterwards. It is an implementation guide, so some account vocabulary is assumed. If you are still deciding whether to run the campaign type at all, start with our article on Performance Max for healthcare lead generation.

No results are reported below, because the right answer differs by account. The method is the point: you can run it in an afternoon on your own data.

Why brand demand distorts the number

Consider a hypothetical, labelled as such. A multi-specialty hospital in Hyderabad spends a fixed monthly budget on one PMax campaign. Some of the resulting enquiries come from searches such as the hospital's name plus "appointment". Those searchers were already considering the hospital, perhaps after a referral, a relative's recommendation or a previous visit. They convert easily and cheaply. If those enquiries make up a large share of the campaign's total, the blended cost per lead falls, and the campaign appears to be acquiring patients efficiently when much of it is harvesting existing intent.

Two decisions follow from that. First, where the brand audience matters to the hospital, it should be captured deliberately, usually by a branded Search campaign with its own budget and its own target. Second, the PMax campaign should be measured on what is left: demand from people who had not named you. Only then can you ask whether the campaign earns its budget.

What Google's brand exclusions do

Google's help page on brand settings for Search and Performance Max explains the tool. Brand lists are created at the account level and applied to individual campaigns. Exclusions stop a campaign serving on queries associated with the listed brands, and are available for Search and Performance Max. For Performance Max, the page states that exclusions apply to Search and Shopping inventory. Where a campaign has both inclusions and exclusions, exclusions take priority.

Three details from that page change how you use it.

• Brand lists use the newest version of broad match, so they match a brand across languages without you entering each spelling. For a hospital with an English name, a Hindi transliteration and an Arabic spelling, that helps, but you should still test it on real queries.
• Variants that share no common word must each be listed. If your hospital has a trading name that is unrelated to the legal name, add both.
• Brand settings limit traffic, and Google notes performance may decrease. Apply them only where a campaign needs them.

Brand exclusions are about Search and Shopping inventory. Display, YouTube and other surfaces inside PMax are not controlled by this setting, so brand-related exposure there needs a different check.

Brand list versus account-level negatives

Google also offers account-level negative keywords, which apply across campaign types, including Performance Max, for Search and Shopping. They are a blunter instrument. A negative at account level applies to every campaign it covers, including the branded Search campaign you want to keep. If you add your hospital's name as an account-level negative, you may silence your own brand campaign. I could not confirm current numeric limits on account-level negatives from Google's own pages, so check the limit in your account before planning a long list.

A workable division of labour:

Build the split in five steps

1. Collect the brand vocabulary. List the legal name, trading name, doctor-led clinic names, branch names, common misspellings and transliterations. Ask the call-centre team which names callers actually use.
2. Create one brand list in the account and add the hospital. Check whether Google's brand library already recognises it, and request addition if it does not.
3. Apply the list as an exclusion to the PMax campaign. Record the date; the date is your before-and-after marker.
4. Make sure branded demand has a home. Run a Search campaign for the hospital's own name with its own budget, so people searching for you still find you at the top. Exclusion without a home for brand simply loses those enquiries.
5. Tag the two streams in your lead log. Add a field for the campaign that generated each enquiry, so the CRM can compare brand and non-brand enquiries on booked and completed appointments.

Measure the before and after properly

The search terms or insights reports in Google Ads show which query categories PMax served on. Before you change anything, export a period of search term data and mark every row that contains a brand term from your vocabulary list. Count the enquiries and the spend attached to brand and non-brand rows. That is your baseline. After applying exclusions, wait for a full conversion delay window; Google's help on conversion delay explains that conversions can be reported up to 90 days after a click and recommends ending report date ranges at least 30 days back, longer for longer windows. Comparing a week before with a week after will mislead you.

Expect three things, none of which can be quantified in advance for your account.

• The PMax campaign's reported conversions may fall, because the cheap brand conversions have left.
• Its cost per conversion may rise, which is the honest number.
• Total hospital enquiries may stay flat if the brand Search campaign picks up the brand traffic, or fall if it does not. This is the number the CMO should watch.

If total enquiries drop and stay down, the exclusion removed demand you were relying on, and the branded campaign needs more budget or better coverage.

The edge cases hospitals run into

• Doctor names. A surgeon's name is a brand of its own, and patients search it directly. Decide whether the doctor's name belongs on the hospital brand list, in a doctor-specific Search campaign, or both. Check that any doctor advertising complies with the NMC rules on professional advertising; our NMC compliance handbook covers the framework, and your compliance owner should review wording.
• Multiple brands in a group. A chain with several hospital brands needs one list per brand, or one list per market, so you can apply the right one to each campaign.
• Competitor names. Brand exclusions can also keep PMax from serving on competitor names, which many teams prefer. Check Google's trademark and policy documentation before bidding on competitor names anywhere.
• Branded queries on non-search surfaces. YouTube and Display exposure to people who already know you is harder to separate. Treat it as a known limit of the report, and compare total enquiry trends rather than relying on platform attribution.

A worked reading, with invented numbers

To make the arithmetic concrete, take a hypothetical account. The numbers are invented for illustration and are not a benchmark. Before the exclusion, PMax reports a set number of enquiries at a blended cost per enquiry. Your marked search term export shows that a meaningful share of those enquiries came from rows containing the hospital's name. Split the spend and enquiries between brand and non-brand rows, and recompute cost per enquiry for each. The non-brand figure is almost always higher than the blended one, because brand rows pull the average down.

Now carry that forward. If the business target is a cost per qualified enquiry for new patients, compare it with the non-brand figure, not the blended one. If the non-brand figure misses the target, the campaign needs work on assets, landing pages or signals, and the earlier blended number was hiding that. If it meets the target, you have a defensible case for the budget. Either way, the decision now rests on the right number.

Keep the brand search campaign on a separate lens. Its job is to make sure people who type your name find the right page quickly, and to defend the name against competitors bidding on it. Measure it on coverage and on completed appointments, and do not claim its enquiries as proof that paid media is creating new patients.

What a good report looks like afterwards

Report two lines side by side every month: new-patient demand (PMax and non-brand Search) and existing-patient or brand demand (branded Search). Under each, show spend, enquiries, booked appointments and completed appointments. Resist adding them together. The two lines answer different questions, and an averaged number answers neither.

If your enquiries land in different places (calls, forms, WhatsApp), a single enquiry log helps. A CRM such as Pioneer CRM is one place for enquiries, lead ownership and follow-ups, which makes the brand and non-brand comparison on booked appointments possible rather than guesswork.

Where to start

1. Export search term data for the last full period and mark brand rows.
2. Write the brand vocabulary list, including transliterations and doctor names.
3. Create and apply the brand list as an exclusion on PMax.
4. Confirm a branded Search campaign exists and is funded.
5. Add a campaign source field to the lead log.
6. Review after a full conversion delay window, comparing total enquiries and not only PMax cost per lead.

Our Performance Max management service includes this separation as part of account setup, and the wider Google Ads guide for healthcare sets out where PMax sits next to Search.