If you missed my blog on the best practices for setting up Performance Max (PMax) campaigns, take a read through that first. Now, let’s talk about what to do after you’ve launched your PMax campaign.
We’re seeing impressive results from PMax campaigns almost right off the bat. I know a lot of people are concerned that it’s just cannibalizing the existing campaigns. And it definitely will reduce the volume standard Shopping campaigns are getting (though you will still get some traffic), and Smart Shopping will be completely stopped. But we’ve seen little to no effect on existing video, display, or Discovery campaigns.
I know I mentioned this in my prior blog, but I feel it’s worth repeating: we create a negative keyword list with brand terms in it and then have our Google rep add it to the account so we can try to prevent conversions just moving from our brand campaigns to the PMax campaign.
A note about negative keyword lists: The good part is once a negative keyword list is added to a campaign, you can keep adding to it; the bad news is that every time you create a new PMax campaign, you have to ask your Google rep to add that brand negative keyword list to the campaign. (My plea to Google – please allow us to do this without Google rep intervention.)
How do we evaluate performance? Mostly we’ve been looking at the overall performance of the account and comparing the account’s results before launch. We also look at traffic and conversions to that segment. Since we usually mirror segmented Shopping and Search campaign structure (for example, we’ll have a Saddles Search and Saddles Shopping campaign, and then we’ll build a PMax Saddles campaign), we’re able to use filters to see what the overall lift has been.
Of course, it’s not perfectly scientific – there could be other factors at play like seasonality. We’ve also been launching campaigns one by one in the same account, so that makes the overall account data harder to evaluate. But for the most part, it works well enough.
Performance Max Campaign Optimization
Time to talk optimization. While there’s not a huge amount of optimization you can do in a Performance Max campaign, there’s a lot more here than you might think, particularly if you’ve been using Smart Shopping.
The Insights tab can be pretty handy once you’ve let the PMax campaign build up some data. The default is Week over Week, so I usually change to Month over Month as soon as I can for more data.
- Consumer Interest – Search Terms: I like looking at this list to see how good of a job (or not) Google is doing with matching. I cycle through sorting by impressions, then clicks and conversions. If poorly matched terms are showing up high for impressions or conversions with no corresponding conversions, I have added a second negative keyword list for that campaign (and, of course, I have to ask our Google reps to add it).
- Audience Insights: some PMax campaigns will have Audience Insights which will show the top-performing audience segments. While not directly useful for the existing Performance Max campaign, unless, of course, the audience is so far off that it gives you clues to update your Audience Segment. But those audiences might be useful to add to other campaigns as a layering tool.
Another way to optimize is with your assets. Looking at the Performance column, you’ll be able to see Low, Good, or Best. Try to update the Low performers. (We see the Short Description and the 960×1200 images tend to have lower results, or at most Good. Not sure why.)
If you are using a product feed, you’ll then want to take a look at the Products tab. You can sort by Cost to see what’s spending the most and sort by Conversions to see what’s converting. If what’s spending money is not converting, you can consider excluding it from the Listing Group. We’ve even tested PMax campaigns with only one SKU, and shockingly, they’ve worked better than expected. One note about the Listing Group is they cap out at 1,000 SKUs, so you won’t be able to see more than that. It’s why the Product page is a better choice for evaluating if you have a huge number of SKUs.
A Word of Caution About Budgets
Let’s talk next about budgets. If you have a PMax campaign doing well that you want to increase the budget for, don’t get crazy with your increases. We’ve seen multiple instances where getting too aggressive with budget changes tends to tank campaign performance, but stepping your budget up slowly seems to work better at maintaining the same ROAS. We assume that a drastic budget increase forces Google to go after searchers much higher in the funnel at too fast a rate, but stepping up the budget makes it smarter at finding converters.
A Girl Can Dream
And finally, here are some things you can’t do with a Performance Max campaign that I’m hoping Google takes heed of and adds in the future:
- Promotional text that’s schedulable: because there are no individual ads, you can’t schedule sales and promotions that start or end outside of working hours. Everything has to be updated manually. Having Promotional ad text that you can schedule outside of the main copy would be hugely beneficial because while you can add Promotion extensions, they don’t allow more detail, nor can you do free shipping promotion extensions.
- Automated rules for asset groups: you can’t use rules to pause or enable an Asset Group. This is one more reason why you should have only one asset group per campaign, so at least you can use a rule to pause or enable a campaign.
- Opt-out options for brand safety: I know there are some issues for clients that have brand constraints or need more control over where an ad shows, so PMax is never going to be a viable solution for them until Google allows for more control. That includes negative keywords that are easier to add, opting out of different channels, etc.
While Performance Max campaigns still leave a lot to be desired from a transparency standpoint as well as having some issues that I hope Google will address, and we don’t know how they will continue to perform for the long term, there’s a lot of potential here. PMax campaigns won’t be a fit for every client, but they are worth testing, and of course, they will be the only option for anyone who is currently running Smart Shopping or Local campaigns.