TPE #92: You can’t trust pMax data (3 simple steps to fix it)

Read time: 8 minutes

Hey, Miles here with The PPC Edge.

You can’t trust your Performance Max results.

Why? Because they’re heavily inflated by Branded Search conversions.

Since Google launched Performance Max in 2021, I’ve been super vocal about why I think you should exclude branded traffic.

But a lot has changed since I first shared my thoughts about this 1.5 years ago…

So in today’s issue of The PPC Edge, I’m giving you an updated guide on why and how to exclude branded traffic from pMax (and why you need a Branded Standard Shopping campaign to prevent your ads from not showing on the Shopping network).

Let’s dive in!

 

Why you can’t trust your pMax data, and why you should exclude brand.

The biggest problem with Performance Max: the data is inflated with branded search conversions, making it look better than it actually is.

Performance Max heavily cannibalizes your other campaigns and LOVES to take credit for every single conversion. People who search for your brand would’ve converted anyway, so why should you let pMax take credit for it?

On average, we see 10-20% of conversions on pMax are branded. That’s a lot!

Even with tight Branded Search campaigns, misspellings will slip through.

If you have a big share of branded search conversion in Performance Max, you simply cannot trust the data. This will lead to wrong decisions, and your account as a whole becomes less scalable.

Let’s back up this claim with real data.

 

 

Make your pMax data more trustworthy so you can scale better.

Here’s the simple 3-step process to make your pMax data more trustworthy (making it easier for you to scale confidently):

1. Analyze your share of brand in pMax.
2. Exclude brand from pMax (2 methods).
3. Catch brand traffic in Branded Search and Branded Shopping campaigns.

Let’s zoom in on each step.

 

1: Analyze your share of brand in pMax.

Analyzing your share of brand traffic in Performance Max used to be a bit of a manual process. Luckily for us, the team at Smarter Ecommerce (SMEC) have built an amazing script that allows you to see the share of brand within pMax over time.

Here’s an example from one of my accounts with <0.2% of brand conversions in pMax (that’s a job well done).

 

 

But I’ve also seen extreme cases. Here’s an example from Denie Geertzen, who audited a new account and found out that over 80% of pMax conversions were in fact branded (!!!):

 

 

That’s truly insane. Just imagine how much you’re overpaying for conversions and clicks…

Listen, 80% is very extreme — but on average I see 10-20% branded conversions with Performance Max.

It would be much better to exclude brand from pMax, and catch your branded traffic in dedicated Branded Search and Branded Shopping campaigns.

If you follow this process, this will happen:

• Your pMax ROAS/CPA will get worse, but it’s much more reliable which makes it easier to scale (because it doesn’t contain branded searches that were going to convert anyway).
• You can maximize visibility (impression share) on branded searches with your Branded Search and Branded Standard Shopping campaigns.

So the next question is: how do you exclude branded traffic from pMax? Let me show you.

 

2: Exclude brand from pMax (2 methods).

You can easily exclude branded traffic from pMax with 2 different methods:

1. Brand Exclusions
2. Negative Keywords (Lists)

 

 

Can’t decide which one you should use?

• Only want to exclude Search and Shopping traffic? → Use Brand Exclusions.
• Also want to exclude Video and Display traffic related to your brand? → Use negative keyword lists.

Brand Exclusions are a great feature, but I’ve seen more success with Negative Keyword Lists (it’s much faster and tends to be more accurate).

I’ll show you how to use both methods so you can test and see what works best for yourself.

 

Option 1: exclude brand from pMax with Brand Exclusions.

Here’s how to set up Brand Exclusions in Google Ads (4 simple steps):

 

Step 1: In your pMax campaign, go to settings.

Click on a pMax campaign, go to settings, and scroll down until you see ‘Brand Exclusions’.

 

 

Step 2: Create a new brand list.

You can exclude brands by creating a brand exclusions list directly inside your pMax campaign, or in the Shared Library.

Enter a list name, and add brands to your list:

 

 

If you don’t see your brand name, you can request it to be added. This process typically takes 4-6 weeks to complete, so be mindful of that.

 

 

Step 3: hit save and make sure the brand list is added to the right pMax campaign

After you save the list, make sure you add it to the right pMax campaign(s). You can create different lists for different campaigns, or use one across all of them.

 

 

Step 4: check if you no longer see branded search terms in pMax campaigns.

The last step is often overlooked: after adding Brand Exclusions, use the Brand Analyzer script to check if you no longer catch brand traffic in pMax.

Pro tip: periodically (every 7-30 days) check if you see any branded search terms in pMax via insights > consumer spotlight > search term insights.

 

 

Option 2: exclude brand from pMax with Negative Keyword Lists.

My preferred way to exclude brand from pMax is by using Negative Keyword Lists.

Back in the day, you needed to ask a Google Support Rep to add your Negative Keyword Lists.

Today, this process is much easier.

Simply fill out the “Performance Max Campaign Modification Request Form” and a rep will add your Negative Keyword List for you.

Use the following selections:

 

 

At the bottom, you’ll be able to download the “pMAX: Request Template”.

 

Fill it out with your desired Negative Keyword Lists and/or negative keywords, campaigns, and match types.

A Google Support Rep will now add them to your campaigns!

 

 

3: Catch brand traffic in Branded Search and Branded Shopping campaigns.

Ok, you’ve now excluded brand traffic from your Performance Max campaigns.

But the job is not done yet!

You need to ensure you catch your brand traffic elsewhere:

• Search → Branded Search campaign.
• Shopping → Branded Shopping campaign.

This is what it looks like:

 

 

Most of you will already have a Branded Search campaign, but you may not necessarily have a Branded Shopping campaign yet.

 

 

If you follow this process, this will happen:

• Your pMax ROAS/CPA will get worse, but it’s much more reliable which makes it easier to scale (because it doesn’t contain branded searches that were going to convert anyway).
• You can maximize visibility (impression share) on branded searches on Google Shopping with a Branded Standard Shopping campaign.

Look at the screenshot below from one of my accounts (it’s a bit small so I’ll explain it in more detail below):

 

 

You see 1x Branded Standard Shopping campaign.

ROAS (conv. value / cost) = 33.64, Search Impression Share = 94.43%. The latter is a huge reason why Branded Shopping is so crucial: you actually KNOW how visible you are for branded searches on the Shopping network by looking at the Search Impression Share.

With pMax, you don’t get this metric, so you have no idea how well you’re defending your brand. Sure, Auction Insights are useable, but it’s aggregated data of both branded and non-branded traffic.

 

You also see 2x Non-branded Performance Max campaigns.

ROAS (conv. value / cost) = 2.47 and 2.65. Much lower, but very reliable to scale because all searches are not brand related (tROAS is 250%).

Before we dig deeper, let’s look at 2 common misconceptions people have about this setup:

 

Misconception 1: “Performance Max will take priority over Standard Shopping so this setup won’t work”.

Well, it’s true that Performance Max takes priority over Standard Shopping if you run them alongside each other with the same products.

However, if you exclude branded search terms from pMax (step 2), there will be nothing to take priority over (because you excluded the search terms!).

 

Misconception 2: if you exclude branded search terms from Performance Max, you’re taking away valuable conversion signals.

I agree: we don’t want to limit pMax by taking away valuable conversion signals. However, there won’t be a problem because the conversion data is shared across campaigns when you use data-driven attribution.

Alright, so now on to how to actually make this work.

 

How to set up a Branded Standard Shopping campaign.

Now it’s time to set up your Branded Standard Shopping campaign. I won’t show you every exact step (speaks for itself), but take in mind these best practices:

• 1 campaign is enough, don’t overcomplicate it.
• 1 ad group could be enough, unless you want to segment by products/categories, that’s also fine.
• Create multiple product groups either split up by product category or product ID. This makes it easier to manage the CPCs of specific products/categories.
• Select the right feed and location targeting.
• Use Manual CPC unless you have a specific ROAS target for branded searches, then use tROAS (I like Manual CPC for branded traffic to keep my CPCs low!).
• Set campaign priority to low

Congrats! You now have a Branded Standard Shopping campaign.

But… You’re not there yet. There is one crucial step you must take to ensure your Standard Shopping campaign actually only catches branded search terms.

 

Pro tip: exclude non-branded search terms from Branded Shopping.

There is one big problem: “Branded Shopping” doesn’t really exist. You can’t target specific (branded) keywords. With Google Shopping, you are automatically matched to relevant search queries based on your product feed.

Therefore, you will be matched both on branded, AND non-branded search terms.

That’s why you must exclude every non-branded search term you see in the search term report of your Branded Shopping campaign.

 

Here’s how I do that:

1. Create a negative keyword list.
2. Add the negatives to your Branded Shopping campaign.
3. Go to your search term report and create a filter: “search term does not include [branded search term + ALL misspellings]”, “added/excluded: none” (you should now only see all non-branded search terms that were matched on Branded Shopping)
4. Select all search terms and select “add as negative keyword”.
5. Click “negative keyword list” > select the list you created in step 1 (add all as exact match negatives).
6. Save this filter so you can easily use it in the future.
7. For the first few days, add all the non-branded search terms that were matched. Then, add it to your weekly checks.

 

I recommend to use a negative keyword list because you can add it to multiple Branded Shopping campaigns (if you have them). This is especially relevant and easy if you have campaigns with the same language in different countries.

It’s a bit of a manual process, but if you do this for a few weeks, your campaign will be almost completely branded. As time passes, the amount of search terms you need to exclude every week will get smaller and smaller.

 

 

And that’s how you make your Performance Max campaigns more scalable!

So that’s the simple 3-step process to make your pMax data more trustworthy (making it easier for you to scale confidently):

1. Analyze your share of brand in pMax.
2. Exclude brand from pMax (2 methods).
3. Catch brand traffic in Branded Search and Branded Shopping campaigns.

Give it a try, it works really well.

Thanks for reading — and see you next week!

Cheers,

Miles (& Bob)