Performance Max in 2026: When to Use It or Avoid It
Should you use Google Performance Max? When PMax fits your account, when Search does a better job, and how to test it without wasting budget.
How to set up Performance Max in 2026: campaign structure, feed fixes, profit values, exclusions and the reports that show which asset group makes money.

This post covers how to set up and control Performance Max. If you haven’t decided whether PMax belongs in your account yet, start with when to use Performance Max and when to avoid it.
Performance Max is one campaign that serves across YouTube, Display, Search, Discover, Gmail and Maps, plus Shopping when you connect a product feed, with one budget and one bidding goal (Google Ads Help). It optimizes for the conversion action you choose, at the value you report back.
That value decides everything. PMax does what you tell it. Feed it revenue and it chases revenue. Feed it profit and it chases profit. Feed it every form fill on the site and it will buy you a pile of junk leads.
You can’t bid per keyword or hand-pick every placement. You still control the three things that decide the outcome:
A simple pattern for ecommerce accounts:
For lead generation, replace margin tiers with lead value tiers and get offline conversion imports running before you launch. Our Google Ads for lead generation page covers that setup.
In Shopping-heavy PMax, the product feed does the targeting. Titles, product types and attributes tell Google which queries to match you to.
Feed fixes that pay off:
product_type with the full category path (for example “Home > Women > Dresses > Maxi Dresses”). Google uses it to structure bidding and reporting in your campaigns (Merchant Center Help).custom_label_0 to custom_label_4 for margin tier, stock depth and seasonality. Google built these labels as filters for bidding and reporting in Performance Max and Shopping (Merchant Center Help).A clean feed usually moves performance more than a batch of new video assets. For more on feed and campaign structure for online stores, see our ecommerce PPC services.
The highest-impact change in most ecommerce accounts is switching the conversion value from revenue to profit. With ProfitMetrics or a server-side value adjustment, the value you send to Google becomes gross profit after cost of goods, shipping, payment fees and returns. Google also offers native gross profit reporting if you send conversions with cart data and add cost of goods sold to your Merchant Center feed.
The first weeks feel uncomfortable. Reported ROAS drops. Profit rises, because the algorithm stops buying high-revenue, low-margin orders to hit a target that never reflected reality. Target ROAS vs profit bidding walks through the switch and how to set the new target.
Set your target once you have two to four weeks of profit-based data, not before. Getting the values right is a tracking job before it’s a bidding job.
Before a PMax campaign gets meaningful budget, apply:
The location setting alone cleans up a lot of waste in accounts that serve a single country.
PMax used to hide most of what it did. That changed in 2025: Google added a search terms report, channel performance reporting and asset metrics with impressions, clicks and cost (Google).
Build a weekly view, in Google Ads or Looker Studio, that shows:
If you can’t answer “which asset group made money last week?” in under a minute, your reporting isn’t finished.
Performance Max punishes fiddling. Set it up with care, then leave it alone long enough to learn. If you want us to build or audit the setup, see our Google Ads management.
Yes. Google rolled out a search terms report for Performance Max in 2025 with the same detail as Search and Standard Shopping campaigns (Google).
Yes, at account level and at campaign level. They only apply to the Search and Shopping inventory PMax serves on, not to YouTube, Display, Discover, Gmail or Maps (Google Ads Help).
One per distinct audience or product story. More asset groups with the same creative split your data without giving the algorithm anything new to test.
Give a new campaign at least a few weeks without major edits. For a formal test, Google recommends running experiments for at least 4–6 weeks, longer if you have a long conversion delay (Google Ads Help).
We use cookies for analytics and advertising. You can opt out of the sale or sharing of your data at any time.Cookie Policy
Your browser's Global Privacy Control signal is on, so we don't share your data for advertising.