Maximizing Revenue with Performance Max 

By: David Anderson, Paid Search Manager

Google’s Performance Max (“PMax”) is a campaign type within Google Ads that allows advertisers to show their products across different Google networks: Search, Display, Discover, Maps, Gmail, and YouTube.If you are supporting your e-commerce business with paid search, PMax’s ability to extend your reach and use machine learning to optimize campaign performance continually makes it an integral part of your strategy.

For most users of Performance Max, the primary goal is the same as it is for most eCommerce businesses: to generate orders and revenue. However, PMax only has a few apparent levers to pull for revenue optimization, and there isn’t much visibility into asset performance. Most PMax best practices are centered around setting-up asset content, asset group structure, granularity, product feed curation, etc. But there are some active optimization levers to pull, and one of those is using a Maximum Conversion Value bidding strategy incorporating targeted Return on Ad Spend (tROAS).

PMax campaigns offer two types of bidding strategy: you can optimize for Conversions or Conversion value. If your primary goal is revenue rather than achieving the highest number of orders, you will select the latter. However, this selection on its own still offers minimal opportunity to optimize. Fortunately, the option to include tROAS adds a useful optimization lever. You can maximize revenue by optimizing the ROAS percentage target to find the best balance between ROAS and Average Order Value for the products in your campaign.

The PMax algorithm prioritizes the products shown and where they are delivered within the Google Network to the right targeted audience, depending, in part, on the tROAS percentage selected. A high ROAS percentage target tends to cause the algorithm to choose “big ticket” items that will return a higher ROAS. Showing these high-value items to the right audience-targeted searchers will drive a higher conversion rate and average order value.


There are multiple strategies around choosing a higher range percentage vs. a lower percentage that tie into performance and scalability. We work with our clients to determine their own optimal ROAS target percentage. Understanding the product mix, particularly regarding the price range, is essential. We model out what kind of proportional sales volume our clients can expect. It’s also essential to adjust the target if products in a particular price range become out of stock or unavailable for other reasons. Because PMax won’t show those products, you will want to have the ROAS target reflect product availability. For our clients, we are seeing ROAS in the ranges of 350% to over 1000%

Something to consider if you use a different source of truth than Google Ads (such as Adobe Analytics) for conversion data: differences in attribution models will cause some discrepancies. The percentage target specified in Google Ads will return a different ROAS in your source of truth reporting; therefore, you’ll need to adjust accordingly. Increasing Google Ads tROAS by c. 150-200% over the actual target ROAS% has helped us offset some discrepancies and meet our monthly ROAS goals.

As with most marketing and media tactics, it’s always a bit of art mixed with science, so it’s still important to consider other PMax best practices, including:

  • Allowing campaigns to run for 45 days to give the algorithm time to learn and optimize performance
  • Adding as many assets as possible to the limit for each type
  • Using relevant audience lists as signals aid the algorithm learning process in more
    accurately targeting each searcher.
  • Creating as many asset groups as possible with customized assets and audiences