Performance Max

Automation has become an increasingly significant part of paid media over the past few years, and as much as we’ve seen great results from the likes of Smart Shopping, RSA and Smart Bidding, for someone like myself who’s worked in paid media for over a decade the diminishing amount of data we are able to see does irk me somewhat and can make me a little resistant to change.

But change is afoot.

Last year Google Ads introduced Performance Max, a campaign format that incorporates all aspects of Google’s ad formats from search ads to display to video ads. So one campaign can cover all aspects of the marketing funnel using themed assets, this is combined with high-intent audiences which signal to Google who to serve the variety of ads to, to drive your chosen conversion point.


“A single-channel approach to bidding limits your ability to maximize performance. Automation can help you find the next highest-value opportunity, no matter where the customer is spending time. This helps drive more conversions within your budget and is known as “marginal cost optimization”. 


The New Customer Acquisition Goal, which can be utilised in Smart Shopping campaigns, has also recently been rolled out to Performance Max to add another relevant signal to Google.

And now retailers can also include their shopping feed into the mix with the Performance Max for Retail campaigns. But it doesn’t stop there, from Q3 Google will begin to automatically update all existing Smart Shopping campaigns to Performance Max, so therefore any resistance to change is likely to be futile.

We’ve started launching Performance Max to begin gathering insights from Retail campaigns. From our initial tests, two out of three clients’ performance was worse with PMax than the previous campaigns we were running. We experienced similar results when Smart Shopping was first introduced so it may just need time for the algorithm to learn before we see strong or stronger performance with PMax

To date we’ve seen stronger ROAS and lower CPAs when comparing against search and shopping activity, we’ve yet to see any new customers convert through Performance Max, but it’s early days in the attribution window.



  • Set up is easy as you just add any assets and some copy and Google does the rest via testing
  • It’s easier to manage one campaign that can target Search, Display, Shopping and YouTube than multiple campaigns (PMax will take priority over all other campaigns)
  • Machine learning does a lot of the optimisations that would usually be done manually


  • Lack of control and data insights
  • Even when the goal is set up to maximise revenue we found the campaigns pushing traffic to irrelevant pages even after we tried to block these pages
  • If performance isn’t as good as in previous campaigns, there isn’t much you can do about it


Performance Max is unchartered territory that sucks in more data points but gives us less data to decipher Google’s black box. We, as data-hungry folk, are always looking for more ways to adjust our optimisation levers and view performance but there are definitely opportunities for Performance Max to be tested and see just how this incredibly simple, and unoptimisable way of running a campaign will really work against our non-PM campaigns.