# Branded Search Cannibalization and PMax Brand Exclusions

> How to measure true incremental value of branded search ads, set up PMax brand exclusions properly, and right-size brand campaign budget based on causality, not reported ROAS.


**By Murtaza Rangwala** · **Published:** Jun 09, 2026 · **Read time:** 8 min read · **Category:** Performance Max

The branded search debate has been going on for as long as Google Ads has existed. *"Why are we paying to appear next to our own brand name when we'd rank #1 organically anyway?"* It's the question every CFO eventually asks, and most marketers give the same scripted answer: *"competitors might bid on our brand,"* or *"we control the messaging,"* or *"the conversion rate is so high it's basically free."*

Some of those answers are partly right. Most of them aren't measured. With Performance Max now in the mix and **brand exclusions** finally available, the question has new dimensions — and new ways to actually measure the answer.

Here's how to think about brand cannibalization, how to use PMax brand exclusions properly, and how to measure the real incremental value of brand spend.

!!!
**The TL;DR:** Branded search ads cannibalize organic traffic at higher rates than most marketers admit — often 60-80% of "branded conversions" would have happened anyway. PMax brand exclusions let you stop PMax from soaking up your brand searches, but only if you have a dedicated brand campaign capturing them. Measure incrementality with geo holdouts, then size brand spend to the real incremental value, not the reported ROAS.
!!!

## What "brand cannibalization" actually means

When someone searches *"Atlantis Dubai tickets"*, they're already looking for you specifically. If your brand search ad appears, they click it. If it doesn't, they click your organic listing. In most cases, the conversion happens either way.

So a "branded conversion" credited to Google Ads is often a conversion you'd have gotten organically for free.

The question isn't *"did the brand campaign generate the conversion?"* — Google Ads says yes. The question is *"would the conversion have happened without the brand campaign?"* In most cases, the answer is also yes.

> Reported brand ROAS is one of the most misleading numbers in performance marketing. It's also one of the highest-looking, which is why nobody questions it.

## When brand search is genuinely incremental

It's not always cannibalization. There are real cases where branded search produces incremental value:

- **Competitor bidding on your brand.** If a competitor is bidding on your brand name, not having a brand ad means they get the click.
- **Brand-plus-modifier queries.** *"Atlantis Dubai tickets discount"* is a different intent than just *"Atlantis Dubai"*. Sitelinks and offers in the ad can convert these better than organic.
- **Mobile real estate.** Mobile SERPs show ads plus map + answer box. Sometimes the organic listing is way down. The ad is the only above-the-fold path.
- **High-intent CTAs.** *"Book Now"* + price extension on an ad can outperform an organic listing for click-to-conversion rate.
- **Brand awareness in new markets.** If you're new in a region, brand searches are rarer and capturing them quickly matters.

But: these cases describe a fraction of total branded query volume. The bulk of brand searches in most accounts are returning customers, people who saw a TV ad, or people doing the final "let me just google the brand name" check. They were going to buy from you regardless.

## How PMax made it worse (and how the fix arrived)

Until 2024, Performance Max would happily serve on your brand searches by default. Even if you had a dedicated brand campaign with manual bids, PMax could still appear on branded queries, double-charging you and inflating the perceived performance of PMax.

This caused two problems:

1. **PMax got credit for conversions** that would have happened on brand search anyway.
2. **Brand campaign performance looked worse** because PMax was taking a share of branded clicks.

In 2024, Google rolled out **account-level brand exclusions for Performance Max**. You can now tell PMax not to serve on specific brand lists (your own brand, competitor brands, anything you want excluded).

This single feature, used properly, fixes one of the biggest measurement headaches PMax accounts have.

## How to set up brand exclusions for PMax

1. In Google Ads, go to **Tools > Shared library > Brand lists**.
2. Create a new brand list.
3. Add your own brand and any variations
4. Save the brand list.
5. Open each Performance Max campaign.
6. Go to **Settings > Brand exclusions**.
7. Apply the brand list.

PMax will now stop serving on those branded queries. Your dedicated brand Search campaign (which you should have) will capture them instead.

> The order of operations matters: don't apply brand exclusions if you don't have a dedicated brand Search campaign running. Otherwise you'll lose brand impressions to competitors.

## How to measure the real incremental value of brand campaigns

This is where most of the debate gets stuck. People assert *"brand is incremental"* without evidence. The right way to know is to test.

### Method 1: Geo holdout

The cleanest test.

1. Pick two matched regions (high historical correlation, see the geo experiments article).
2. **Treatment region:** continue running brand search campaigns as normal.
3. **Control region:** pause brand search campaigns entirely.
4. Run for 4-6 weeks.
5. Measure **total bookings/conversions** in each region (paid + organic + direct combined).
6. Run Causal Impact to estimate the true incremental lift of the brand campaign.

Most accounts I've tested this way find that **60-80% of branded conversions are not incremental**. Organic search captures most of them when the brand ad is off. The true incremental ROAS of brand search is often 150-250%, not the 1,000-2,000% the dashboard shows.

### Method 2: Time-based holdout (less clean)

If you can't run a geo experiment, you can pause brand campaigns for 1-2 weeks and compare total branded volume against the pre-period.

This is less clean because it confounds the test with whatever else is happening in the market that week (seasonality, news, events). But it's better than nothing.

### Method 3: Dayparting test

Pause brand campaigns for specific hours of the day (e.g., 2am-4am when total traffic is lowest) for several weeks. Compare paid+organic+direct conversions in those hours to other hours. Statistical power is lower but the test is invisible to operations.

## How to size brand spend after measuring incrementality

Once you know your true incremental ROAS for brand campaigns, the budget decision changes.

If your reported brand ROAS is 1,400% but incremental ROAS is 220%:

- The brand campaign is **profitable** (220% > 100% ROAS = break-even on revenue).
- It's **less profitable than alternatives** at the same spend level.
- Continue running brand campaigns, but **stop reporting reported ROAS** as the headline metric.
- Set brand budget based on incremental ROAS, not reported ROAS.

The most common outcome from this exercise: brand budget gets reduced by 30-50%. The savings get redistributed to upper-funnel campaigns where incrementality is higher. Total conversions rise. Total spend either stays flat or drops.

## Use case: a regional restaurant group questioning brand search spend

A composite based on patterns I've seen.

A restaurant group with 14 locations was spending £6k/month on brand search ads. Reported ROAS was 2,100%. The CEO was a "scale it forever" advocate. The marketing manager wasn't sure.

We ran a geo holdout. Two regions matched on size and historical correlation:

- **Manchester:** continued running brand search ads.
- **Birmingham:** paused brand search ads entirely for 6 weeks.

**Findings:**

- Manchester branded conversions: roughly the same as baseline.
- Birmingham total bookings: dropped by 4% during the holdout, **not by anywhere near the 100% you'd expect if every brand conversion was incremental**.
- Organic and direct traffic in Birmingham rose during the holdout — picking up most of the demand that brand ads previously absorbed.

Causal Impact analysis: true incremental ROAS of brand search was approximately **310%**, not 2,100%.

The recommendation: reduce brand spend by 40%, keep enough running to protect against competitor bidding and capture high-intent modifiers (*"book", "near me", "menu"*), redirect the savings into a Demand Gen campaign for new restaurant openings.

**Results over 6 months:**

- Brand spend: £6k → £3.6k.
- Brand search bookings: down 6% (the non-incremental ones).
- Total bookings across the account: up 11% (because the redirected budget drove more new customers).
- CEO finally trusted the data and stopped pushing for brand budget increases.

The brand campaign didn't get killed. It got rightsized.

## Common mistakes

- **Reporting brand ROAS as if it's incremental ROAS.** They're radically different numbers. The first is correlational. The second is causal.
- **Pausing brand campaigns entirely.** Almost always wrong. Some incremental value exists, especially around competitor protection.
- **Skipping the brand exclusion setup in PMax.** Lets PMax double-count brand conversions, distorting both PMax performance and brand performance.
- **Not measuring competitor brand bids.** If a competitor is bidding on your brand, brand ad incrementality goes up. Check the auction insights regularly.
- **Treating brand as "free conversions."** They're not free. Even reported brand ROAS includes real spend. Free is only the *organic* listings you'd have without paid.

## Bottom line

Branded search is real spend on traffic that mostly converts whether or not you advertise. The exact percentage that's incremental varies by account, but for most established brands, it's a minority.

- Set up **PMax brand exclusions** so brand traffic flows to a dedicated brand Search campaign, not PMax.
- **Measure true incremental brand ROAS** with a geo holdout, not the dashboard number.
- **Size brand budget to incremental value**, redirecting the rest to higher-incrementality channels.
- **Continue running brand campaigns** at a reduced budget — competitor protection and high-intent modifier capture are still worth something.
- **Stop reporting reported brand ROAS** to stakeholders. It misleads decisions.

The brands that get this right invest less in brand search and more in growth channels. The ones who don't quietly subsidise organic traffic with paid budget and call it performance marketing.

It isn't. It's bookkeeping.

---

**Sources and further reading:**

- [About brand lists and brand exclusions (Google Ads Help)](https://support.google.com/google-ads/answer/14310886)
- [About auction insights (Google Ads Help)](https://support.google.com/google-ads/answer/2579754)
- [Google Ads geo experiments (Google Ads Help)](https://support.google.com/google-ads/answer/9089888)
- [CausalImpact (open-source library by Google)](https://google.github.io/CausalImpact/CausalImpact.html)

---

**Tags:** brand search, pmax brand exclusion

## About the author

Murtaza Rangwala is a senior independent Google Ads consultant. 8 years, 1,900+ campaigns shipped, $250M+ in client revenue generated. Independent practice capped at four concurrent clients.

- More posts: https://www.murtazarangwala.com/blog
- Book a 30-min call: https://calendly.com/murtaza_rangwala/30min
- Free Google Ads audit: https://www.murtazarangwala.com/#audit