Inside Google Ads podcast: Episode 133 - PMax Mistakes

I keep seeing the same mistakes in Performance Max over and over again. And while that's really good for my business as a Google Ads coach who helps you fix those mistakes, it is not good for your business. 

Please stop.

Today, we're going to cover 5 of the top mistakes I see Google Ads practitioners making with Performance Max campaigns. And to be clear, these are not newbies I'm talking about. My clients are people who work at agencies, freelancers, in-house folks who manage millions of dollars a year in spend. 

So if as I'm going through you think, "Oh no, I'm making that mistake." Fear not. You're in good company. It does not mean you don't know what you're doing, but it does mean you should rethink your PMax strategy.

I'm your host, Jyll Saskin Gales. I spent six years working with big brands at Google and now I work for you. 

This is Inside Google Ads, Episode 133: PMax Mistakes.

1. The first mistake I see is treating asset groups like ad groups and they are not the same thing. 

An ad group in a Search campaign or Demand Gen has true targeting. You pick your keywords or your audiences and that is all Google will target. 

In Performance Max, you have audience signals and search themes. These are mere suggestions to the algorithm, which it could target or it could completely ignore. What that means is the reason to have a different asset group isn't to test different signals. Instead, the reason to have a different asset group is to test different assets. 

So often I see a PMax campaign, we go into it, there's 10 asset groups, exact same text, image, videos in every single one with different signals, and that is backwards. Instead, you could have no signals or sure, give audience signals, and search themes, but every asset group should have different assets. If you're advertising products, it means you're going to want different products likely in different asset groups or different use cases. 

For example, I have a client who used PMax to advertise the same product, but to very different kinds of audiences. And to be clear, that didn't mean we just changed the audience signals. We did that, but also we changed the assets. In one area where we were trying to target a luxury shopper, the images we used, the videos, and the text really conveyed the luxuriousness of the product. 

Whereas, in another asset group where we were really trying to appeal to pet owners, the assets all changed. We showed photos of pets with the product. Our text assets talked about pets, etc. So while we did give different audience signals, the reason this strategy worked is because we gave each asset group different assets which therefore would appeal to different audiences and would cause PMax to show these to different audiences. 

Remember that in PMax your creative is your targeting. It must specifically appeal to your target audience and just as importantly not appeal to your not target audience. 

As Google Ads practitioners, especially if you've been doing this for a while, we've been trained to think of creative as an afterthought. And while that's no longer the case in any campaign type, it is especially not the case in an AI powered campaign like PMax. Your creative needs to be front and center because that is your targeting.

2. Alright, the second mistake I'm seeing a lot of people make, and this may be a controversial one, is immediately excluding your brand from PMax. 

And here's the thing, it's not necessarily a problem if your PMax is targeting brand searches in Search, in Shopping, or both. 

Why? 

Well, especially if we're starting a brand new PMax campaign or this is a low budget account, the thing you need is conversion data, audience data, data density, and so keeping brand in PMax can actually help it hit conversion thresholds faster so that you can start using things like Target CPA or Target ROAS. 

Also, I have seen many examples across geographies, across industries where for whatever reason, branded searches performed better in PMax than they did in a standalone brand Search campaign or brand Shopping campaign. 

Look, if you're working with a large budget, which I would call mid five figures per month or more on Google Ads, then absolutely you can keep them separate. But anything less than that, it may be in your best interest or your client's best interest to just let PMax cover it all. And perhaps then once you get a lot more data and you're scaling up, consider breaking it out on its own or not. 

To be clear, I'm not saying that no one should ever exclude brand from PMax. What I am saying is that if you always default to launch new PMax, exclude brand right away, I encourage you to rethink that because it may actually get better results for you in PMax than in its own standalone campaign.

3. Mistake number three is launching Performance Max for a lead gen objective without Offline Conversion Tracking.

I can't believe I'm still seeing this mistake. If all you're tracking as a conversion is a call from ad, a click to email you, a click to phone call, a submit lead form button, you are setting yourself up for disaster because you know who really loves clicking buttons? Bots, low-quality leads, all the kind of stuff you don't want PMax to go after. 

Again, to be clear, all lead gen advertisers should be using Offline Conversion Tracking. There's really no excuses anymore. But especially with Performance Max, do not even think of turning that thing on until you are at least tracking qualified leads, if not closed won deals. 

If you're not sure exactly what I mean by Offline Conversion Tracking or you need a little refresher, I have a whole video on that on my YouTube channel. I've included the link in the episode description. 

But long story short, do not launch PMax for a lead gen objective unless your conversions are at least a quality lead if not something further down funnel.

4. Coming to our e-commerce side, mistake four is duplicating products across Performance Max and standard Shopping. 

Why are we still doing this?

Taking your products and then deciding to put them all into your PMax and also into your standard Shopping so that they're just fighting and competing against each other in the auction. 

Here's the analogy that I like to give that maybe will help explain why this is such a bad idea. Let's say you have two children and you want to save some money on school. So, you decide to send one to school one day, the other one to school the next day, and on and on it goes, rotating each day. 

By the end of the school year, what's going to happen? Your first kid will have learned about half what they need to learn. Your second kid will have learned about half what they need to learn. Will either one be ready to advance to the next grade? No! They missed too many days of school. They're both going to have to repeat the year again. 

This is what happens when you send PMax and standard Shopping to constantly compete in the same auctions. Only one can show an ad at the time, which means they're duplicating effort and both trying to learn the same lesson over and over again. 

Whereas, when you have just one child, you send them to school every single day. By the end of the year, they've learned what they need to know to graduate to the next grade. 

Just like if this product only exists in standard Shopping, then whenever it's eligible to serve, it goes in standard Shopping and all the data is there together so it can learn faster. 

There are absolutely reasons why you may want to have Performance Max campaigns and standard Shopping campaigns in the same account. Just make sure you are de-duplicating your products so that each product is eligible in PMax or in standard Shopping, not both.

5. Last but not least, mistake number five, which I consider just as egregious as the one we started off this episode with. Do not over-optimize your Performance Max campaign. 

What do I mean by that? Well, I will hop on calls with people who tell me, "Oh, I launched PMax and I'm having to add hundreds of negative keywords a day." If you're having to add hundreds of negative keywords a day, you're doing it wrong. Or Performance Max is doing it wrong, in which case it's not the right campaign type for you or you haven't set it up properly. 

Same thing. “Oh, I launched PMax and now I'm adding new headlines every week and adding new images.” Stop, stop, stop, stop micromanaging. This applies to all AI-powered campaigns, by the way, but especially for Performance Max. 

So, here are a few practical rules of thumb that I recommend you follow to know when and how to actually optimize PMax. By the way, I have a whole episode about optimizing PMax from a few months ago. I will also include the link to that in the episode description. 

Okay, let's talk about search terms. Do not analyze your search terms in a vacuum. 

If it is day two of your PMax campaign and you are seeing some search terms that you're like, why the hell would PMax show my ads on that? Pause for a moment and think: even if this search term isn't relevant, do I still want to reach the searcher? Like the kind of person who would search for that, is that the right person for me? If so leave it in. 

The only reason to exclude and add a negative keyword to PMax is if you would genuinely never want to advertise on that Search or to that searcher. And in that case, feel free to add the negative. 

But then you need to think about why did PMax put me there in the first place? Because remember, Google doesn't just pluck those queries out of thin air. There are billions of queries to choose from each day. It is choosing the ones that it sees as related to your business as determined by your creative, your landing page, your conversion tracking, your bid strategy, etc. 

So, think about it as well. Why did PMax think that was relevant for me? And that may help you solve the problem, perhaps the way you've set up your campaign. 

In general, if you do want to exclude a search term, I recommend waiting until the search term has at least a 100 clicks on it so you can genuinely see if it performs or not for you. But also keep in mind that by that point, PMax will have also learned. So if it's advertising on a search term and that search term brings no business value for you, then guess what? PMax is going to follow your bid strategy and conversion goals and stop bidding on it. So you don't have to come in with a negative. PMax will take care of it for you. If it's not taking care of it for you, some other issue is going on there.

All right. Now, same for assets. By assets, I mean a headline, a description, an image, a video, a product, etc. So, in general, for PMax or any campaign type, I say don't judge an asset's click-through rate until it has at least 100 clicks and don't judge an asset's conversion rate until it has at least 50 conversions at the asset level, which you can see on your Asset Performance Report. 

So, if you're going to review your asset performance in PMax and you see most of your headlines have 20, 30 clicks, it's too soon. Neither you nor Google actually knows what's working yet. Don't over optimize and try to control your Performance Max campaign. If that's the way you want to manage your campaigns, more power to you, but then don't use PMax. Stick to Search, Shopping, and/or Demand Gen instead, where you actually do have those controls the way they're meant to be used. 

I've said here before that PMax is now a clear box, not a black box, because you have full transparency in PMax the way you do in other campaign types. And something I'm starting to come around to is even more than that PMax gives you practically the same level of control as your standard campaign types. 

If you're comparing PMax to standard Shopping, literally the exact same amount of control and transparency. 

If you're comparing PMax to Demand Gen, the only real difference is audience targeting versus audience signals and channel performance selection. 

Although in PMax, you can now exclude Search Partners if you want. You can exclude the Google Display Network. I'm honestly not sure yet whether I recommend doing that or not. I like to form my opinions based on data. I don't have the data yet, but my point is it's happening. 

And then on the Search side, if you're using Exact Match keywords, we still have a ton of close variants, but like fair, that's kind of control. But if you're using Broad Match keywords or AI Max in Search, again, perfectly same level of control or near similar levels of control in search that you have in PMax. 

So, PMax is not a black box. It's a clear box. And it's a clear box that's actually giving you a lot more control than when it came out. 

In my opinion, this kind of means that the PMax experiment failed. Google wanted us to just give Google, here's my conversion tracking, here's my budget, here's my goal through my bid strategy. It just didn't work the way advertisers needed it to. 

So, a couple years ago, we got all the transparency and the Search Term Report and channel performance and asset performance and all of that. 

And then we got a bunch of the control back like demographic exclusions and placement exclusions and all kinds of stuff. 

PMax can absolutely still be a great vehicle to drive your business, but you need to cut through the misconceptions to truly understand what the campaign does and doesn't do.

To recap, those five mistakes to avoid are:

1. Creating multiple asset groups to test different signals. Instead, the reason to create multiple asset groups is to test different assets. 

2. Do not immediately exclude brand from PMax. Let it start in there and then based on data decide if you want to exclude it or not. 

3. Launching PMax for lead gen without Offline Conversion Tracking. Don't do it. PMax should always be optimizing for at least a quality lead if not a closed won deal, a purchase, etc. 

4. Duplicating products across PMax and standard Shopping. Each product should live in one campaign, not multiple. 

5. Overoptimizing your PMax, by trying to control it rather than letting it learn and do what it needs to do. 

If you're still struggling to get Performance Max to work for you, you can book a call with me on my website at jyll.ca. That's J-Y-L-L dot C-A. 

I'm Jyll Saskin Gales and I'll see you next time inside Google Ads.