Decoding Performance Max

Questions that are answered in this guide:

  • What are Performance Max campaigns?
  • How are Performance Max campaigns an improvement over other campaigns?
  • Setup Related Queries
  • Measurement Related Queries
  • Troubleshooting Related Queries

What are Performance Max campaigns?

Performance Max campaigns are a relatively new campaign type that allows for the targeting of multiple placements, using multiple inventory types. 

Its reason for existing is to supplement Search campaigns by providing greater conversion volumes across GDN, YouTube, Discover, YouTube, and Gmail, while also generating keyword suggestions for Search campaigns.  The impact is to strengthen Google campaigns by having image, text, video and catalog assets work in parallel to generate revenues.

Performance Max is, effectively, the next logical step in the development of Machine-Learning-based campaigns. It combines increased integration of asset types, more advanced insights, and a shift to a more holistic form of conversion reporting.

How are they an improvement?


Performance Max campaigns are an improvement in the following three ways:

1. They Integrate Search, Discovery, Video and Shopping Assets

The creation of Performance Max campaigns mandates the setup of text (headlines, long headlines and descriptions), images (square, landscape, portrait), a product catalog (the same as we use for SmartShopping), with video being optional. 

This allows for those that want to run smaller budget campaigns, but have an abundance of creative assets, to leverage them all simultaneously. This is unlike SmartShopping, where asset addition excluding the catalog is severely limited.

2. Insights on Audiences to be Targeted and Search Terms

We can, at the campaign level, determine which audiences performed well in terms of conversion share, and the Index. 

We can also determine which search terms generated the best conversions, and we can order them by search volume to determine whether they can be used in other ways, such as in the creation of Custom Intent audiences.

This is radically different from all other campaign types, such as SmartShopping (which is notorious for being a black box regarding audiences being targeted) and Search, Display, YouTube and Discovery (all of which only provide information regarding predefined audiences).

3. Internal Reporting Shift from Click-Through Conversions to All Conversions

When we look at the internal reporting in a SmartShopping campaign i.e. the Reporting at the product group level, we see that the data is for “Conversions”, “Conv. Value”, “Conv. Value/Cost”, and so on. This refers only to Click-through Conversions. 

In Performance Max campaigns, we see the results shown in the form of “All Conversions”, “All Conv. Value” and “All Conv. Value/Cost”. This refers to Click-Through Conversions AND View-Through Conversions i.e. conversions that happen when someone sees an ad, does not click, and later converts directly on the site in question.

It is vital that we recognize that revenues are not just a direct function of clicks and cost, but also of impressions, and organic traffic driven by the impressions in question. Part of the reason for monitoring the CRR metric at the overall site level is to make sure that conversions and revenues that may not be directly attributed to AdYogi are also included in the overall picture of business growth.

Setup Related Queries


1. How do I add specific placements in PerformanceMax for my luxury brand/luxury products?

We cannot add placements to be targeted in PerformanceMax. We can, however, exclude specific placements/placement lists at the account level by referring to the placement exclusion report.

2. Is it possible to prioritize a product/products over others within a PerformanceMax campaign?

This is not possible, but we can have a separate campaign targeting a single product/products for the purpose of prioritizing.

3. How do I decide the target ROAS for a new Performance Max campaign?

This is dependent on your ROAS expectation, but we can look at the current Last 30 Days account-level ROAS that you are receiving, or the performance of your current remarketing campaigns, whichever is higher.

4. Is there a way to exclude specific placements (Discovery/YouTube, for example), as a whole, from Performance Max placements?

We cannot exclude any Google owned and operated properties (Google Search, Shopping tab, YouTube Home Feed, Google Discover, Gmail)

5. Why is it not a best practice to use in-market audiences in the Performance Max signals? Wouldn’t adding remarketing data mean directing the campaign to said audiences, and thus we may not be able to get prospecting traffic?

Inmarket audiences do not necessarily equate to the ideal audience. When we use audience signals, we are trying to teach the campaign what the ideal converting audience looks like, so that the machine learning can take effect. The campaign’s actual reach goes beyond the signals. 

Thus, remarketing data would not lead to targeting specific audiences, as much as it would lead to targeting those similar to said audiences.


Measurement Related Queries

1. How can we check performance at an Asset Group Level?

We can use the Listing Groups attached to those Asset Groups as a proxy for general performance, by checking spends, impressions, clicks, conversions etc for all Listing Groups per Asset Group, as a whole.

2. How can we check asset performance in terms of revenue/cost for videos and images?

We cannot see exact data for individual asset performance. However, we can open the Performance Max campaign, go to the Asset Group in question, click on View Details - Assets, and use the report to determine which assets are performing relatively well.

3. Can we get creative-wise metrics/analysis for what is picked up automatically?

We can open the Performance Max campaign, go to the Asset Group in question, click on View Details - Combinations to define the best performing Headline-Description pairs (for text) and Headline-Description-Asset groups (for images).

4. Where can we see audience-level insights?

These are available at the campaign level on the Insights tab, under Audience Insights.

5. Where can we see the split between New and Returning Users on Performance Max?

This can be seen on Google Analytics. Go to Audience > Behaviour > New vs Returning report in Google Analytics.

Troubleshooting Related Queries

1. How can we optimize a Performance Max campaign with a sharp performance drop this month, compared to last month?

We can do the following:

  • Check the Listing Groups, product wise. Has a well-performing product gone out of stock? Have there been any major spend/CTR/conversions changes w.r.t. a particular product?
  • Refresh the audience signals; add more keywords to the custom intent audiences and refresh the first-person remarketing lists (offline purchasers, general customer data, all-time organic purchasers data)
  • Check the Placements Report. Are a large number of impressions going to a particular website/app, relative to the previous period?
  • Check if a particular location has had disproportionate spends, or there has been a disproportionate drop in the conversion rate. Exclude locations/increase midfunnel marketing accordingly.
  • Check Insights tab to see changes in demand/rising queries/falling queries. Optimize these by adding them to your Custom Intent audiences.

2. After working on catalog titles, bidding strategy, and audience signals, what further can we do to optimize performance?

We can add sitelink extensions, edit feed attributes in the Google Merchant Center, add Merchant Promotions and Product Review feeds, both in the Google Merchant Center.

3. How can we analyze the reasons for abrupt overspend in Performance Max campaigns?

We can check which Listing Groups, if any, suddenly increased in spends, and take action. We can also check the Placement Report.