Join 20K+ subscribers

TPE #19: pMax - 7 advanced learnings since July

Nov 28, 2022

Read time: 7 minutes

Since July, we’ve spent over €1.5 million on Performance Max campaigns.

For better or worse…

A lot has changed since then, and we’ve learned a lot.

It’s definitely a hard campaign type to truly master and scale - but not impossible.

In this issue of The PPC Edge, we’ll talk about our 7 biggest learnings since July:

(advanced stuff almost nobody talks about)

 

  1. Why pMax is awesome and trash at the same time
  2. How to fix a pMax campaign when everything goes to sh*t
  3. How to deal with cannibalization of Search campaigns
  4. How to reduce low lead quality for lead-gen
  5. When to choose feed-only vs all assets
  6. When (not) to run additional Display/Video campaigns
  7. How to optimize your assets for conversions

 

Let’s dig in!

 

Learning 1: Why pMax is awesome and trash at the same time

 

Google has truly built something special with Performance Max. One campaign, that can serve ads on all of Google’s inventory.

Amazing!

… For Google?!

We feel like Performance Max is still hit or miss. It can perform extremely well when set up correctly. However, it can also go crazy and spend your budget really fast on low-value placements.

The lack of insights are killing, but here’s what you can do about it:

We use a script by Mike Rhodes, that shows where you’ve spent your budget (Shopping/Video/Other). It’s not 100% foolproof but it will give you some nice additional insights that are otherwise hard to find within Google Ads:

Or you can go next level and use a click-fraud prevention tool like Lunio (they just launched a pMax integration) to stop wasting spend on low-value placements.

The lack of insights are definitely still bothering us, but there’s nothing we can do about it except adapt and try to make the most out of it.

It feel like Performance Max still isn’t a fully finished products, even though a LOT of spend goes through it.

And the limited insights make us wonder whether pMax is truly beneficial for the advertisers… Or for Google…

Anyway, we’re not here to complain. Our job is to adapt quickly and try to make the most out of it.

 

Learning 2: How to fix a pMax campaign when everything goes to sh*t

 

Automation can be tricky. Eventually, something will break. And since Performance Max only has a few levers to pull, it can be hard to identify what to do when this happens.

Let’s look at an example from one of our accounts and what we did to fix it.

In the graph below, you can see a fully filled Performance Max campaign (all assets) that started to spend like crazy. The Target ROAS is 200% (and we were usually getting more than that). After it started to go HAM with the spend, ROAS started to dip.

Also: look at the conversion rate. For about 2 weeks, CVR and ROAS dropped. We tried several optimizations (assets, targets etc.), but nothing really worked. Then we identified the root cause of the problem and were able to fix it (see the huge spike in CVR at the end? ROAS also started to increase gradually).

 

 

Here’s what we did:

In order to fix a ‘broken’ Performance Max campaign, you need to create a diagnosis of what’s going on. Look at yourself as a doctor. When a patient has an issue, the doctor creates a diagnosis and a plan to fix it.

The same applied to us in this situation. We needed to diagnose our pMax campaign.

The problem: spend went up like crazy and ROAS + CVR dropped significantly.

We found the root cause by looking at our spend allocation. We noticed this campaign went all out on spending on video placements. It’s the hardest we’ve seen pMax go on video placements so far.

And we paid for it with a lower conversion rate and ROAS.

After we identified the problem, we were able to come up with a plan to fix it: we decided to pause the asset group with all assets and switched to a feed-only setup.

Back to the basics. No more spend on upper funnel placements, but Shopping/Search only. As a result, CVR and ROAS started to increase again.

How you fix a Performance Max campaign depends on so many factors. The main thing you need to do is create a diagnosis of what could potentially be wrong. Then you implement some changes, wait, see if it actually helped, and readjust accordingly.

Some things to consider:

 

  • Check your spend allocation: where are you spending?
  • Check your budgets: are you spending too much?
  • Check your tracking: is it still working properly?
  • Check your targets: are they too low/high?
  • Check your assets: are they matched to your products?
  • Check your asset groups: should you use all assets or feed only?

 

There is a whole lot more to look at but these will help you on your way to fix ‘broken’ Performance Max campaigns.

 

Learning 3: How to deal with cannibalization of Search campaigns

 

Performance Max campaigns can and will eventually cannibalize your Search campaigns since they basically target the same inventory. Which campaign will win, depends on your keywords and ad rank.

Search will take priority if the user’s search query is identical to an eligible Search keyword of any match type in your account. If queries do not match exactly, the campaign/ad with the highest ad rank will be prioritized.

Most people think you need to add exact match keywords to your Search campaigns in order to get priority over Performance Max. However, it’s more about the actual keyword than it is about the match type.

An example of how matching can work:

 

 

Since there are sooooo many different user queries, Performance Max will eventually take priority over your Search campaigns. It will just happen.

If you want to take back some control, either add more keywords to Search campaigns or simply accept the fact that pMax loves to cannibalize.

You could also work with extensive negative keyword lists especially for pMax but we wouldn’t recommend this.

In the end, the campaign/ad with the highest ad rank will be prioritized which should in theory be beneficial to your performance.

We say in theory, because we simply don’t know. That’s the downside of Performance Max and automation in general.

 

Learning 4: How to reduce low lead quality for lead-gen

 

Lead quality can be a big problem for a lot of advertisers running Performance Max. We've seen a lot of work-arounds - like purposely increasing the amount of form fields - to avoid spam and bot traffic. But these kind of practices have a huge negative impact on the conversion rates of your 'qualified' traffic.

These kind of workarounds are no-go's for us, personally. But what can you do instead?

 

1. Switch over to Offline Conversion tracking

If you are running Performance Max campaigns for Lead Gen clients, tracking offline conversions is a must. You absolutely want to see the amount of closed deals from your Performance Max campaign. With this information you can calculate your lead > sale ratio, CAC, and more. But the most important part is using these datapoints to feed the algorithm with valuable data where your sales are actually coming from.

By switching your primary conversion from leads to sales, you can tell the algorithm to optimize towards generating as much as closed deals as possible (within your target CPA). Keep in mind the conversion volume though. In our opinion you need at least 10 conversions (closed deals) per month per performance max campaign in order to leverage the algorithm fully.

 

2. Use a tool to prevent click fraud and spammy traffic

It’s time to stop paying for clicks that will never convert, and automatically eliminate bots and fake users across Google Ads. You can do this with a tool like Lunio. They recently improved their click fraud software so it works with Performance Max smoothly.

 

Learning 5: when to choose feed-only vs all assets

 

We’re still big fans of using feed-only Performance Max campaigns. This is basically Smart Shopping + Dynamic Search Ads.

Here’s some considerations on when to choose feed-only vs all assets.

 

When to go for feed-only Performance Max campaigns:

  • You don’t have high quality image and video assets
  • You don’t want to spend on Video/Display placements
  • You want to focus on bottom-of-funnel placements

 

When to go for fully filled Performance Max campaigns:

  • You have high quality image and video assets
  • You are comfortable spending on Video/Display placements
  • You want Performance Max to do everything for you
  • Your entire account is performing well and you want to expand

 

We’ve seen both setups work well, but also seen them both not work well.

As simple as it may sound, it’s always a matter of testing.

 

Learning 6: When (not) to run additional Display/Video campaigns

 

This learning is sweet and short: in most cases it doesn’t hurt to run additional Display and/or Video campaigns next to Performance Max. Here’s why:

  1. Typically, pMax spend on these placements is relatively low (1-10%)
  2. Reach on GDN/YouTube is virtually limitless

 

If you are an advanced advertiser, it’s definitely recommended to run additional Display/Video campaigns because it allows you to scale much harder. You will have complete control over your audience targeting, and you can control your spend (which you can’t with pMax).

Even though pMax does some retargeting, we always advise to at least run remarketing campaigns on Display, Discovery and YouTube if you have the budget and creatives.

 

Learning 7: How to optimize your assets for conversions

 

With pMax assets, there are 3 major types to look at:

  1. Image assets
  2. Video assets
  3. Ad copy assets

Let’s look at some learnings for every asset type.

 

Optimizing image assets
  • Always use your own images, try to stay way from stock photos
  • Use high quality images
  • Test images with text vs no text
  • If you use text, make it BIG, otherwise it’s hard to read
  • Keep the main focus on your product
  • Test lifestyle images vs. product images
  • Don’t use your product feed images

 

Optimizing video assets
  • Focus on the hook (the most important part of your video)
  • Introduce your brand early (simple image overlay works)
  • Hold attention with emotion and pacing
  • Use fast and snappy clips
  • Finish strong with a clear call to action
  • Product-focused videos work well for us
  • Build for small screens

 

Bonus tip: preview your videos on mobile, not on desktop!

 

Optimizing ad copy assets
  • Speak to a user’s problems and desires
  • Focus on benefits, not product features
  • Tie your copy to your keywords and products
  • Focus on your headlines (the most important asset)
  • Add as many different items as you can
  • Avoid generic language in your ads
  • Be creative: don’t only use copy from your Search ads

 

TLDR recap

 

Performance Max is here to stay. Our 7 biggest learnings from the last 4 months:

 

  1. Why pMax is awesome and trash at the same time
  2. How to fix a pMax campaign when everything goes to sh*t
  3. How to deal with cannibalization of Search campaigns
  4. How to reduce low lead quality for lead-gen
  5. When to choose feed-only vs all assets
  6. When (not) to run additional Display/Video campaigns
  7. How to optimize your assets for conversions

 

We’re excited to see what next year will bring. We don’t know what will come, but one thing is for sure: Google is going to automate even more.

And we’re here to help you get the best possible performance.

See you next week!

Cheers,

Bob & Miles

โ€‹



And whenever you're ready, here's how we can help you further:

ย 

1. Join our global community of world-class Google Ads Specialists:
Stay up to date with all developments, get trained by us all year round, and learn & grow together with like-minded PPCers.
Become a member of The PPC Hubย (1.400+ members). //ย Enrollment is closed. Join the Interest List to be the first to get notified when the doors open again.


2.ย New episodes of The PPC Mastery Podcast air every Wednesday.
Subscribe now onย YouTubeโ€‹, โ€‹Spotifyโ€‹, and โ€‹Appleโ€‹ย so you don't miss anything!