E-commerce
October 28, 2026
Wondering what new Performance Max features to test in 2026? Three updates deserve your attention: comparative experiences, data exclusions on audiences, and product overlap detection. When used well, they finally make your Performance Max campaigns readable and manageable, instead of a black box that absorbs your budget.
The trap is believing that simply activating everything is enough. These tools are mainly for comparing, deciding, and excluding what does not convert. Without a method, you add complexity without real gains in sales.
So what new Performance Max features to test in 2026? On the agenda:
How do Performance Max experiences work?
Can you compare a Shopping campaign to Performance Max?
What is the purpose of the experience on the final URL extension?
How to measure the real contribution of Performance Max?
Why exclude audiences with your data?
Let's go.
Summary
Why does Performance Max remain a priority for Google Ads?
Performance Max is not just another campaign. Google makes it a central format of its advertising ecosystem, and thus a priority for advertisers. In other words, updates to this format arrive quickly and concretely change how you manage it.
For an e-commerce merchant, the stake is simple: this type of campaign combines Search, Shopping, Display, YouTube, and Discover in a single automated mechanism. You entrust the algorithm with finding the right placements and the right audiences.
The downside is the lack of visibility. You see sales coming in, but you do not understand where they come from. This is exactly what the three 2026 updates correct: experiences to compare, exclusions to filter, and a product report to spot duplicates.
To situate these campaigns within a broader strategy, rely on a solid foundation on the Google Shopping side.

Convert over 2,000 customers on average per month with Qstomy.
The world’s 1st Shopify AI dedicated to customer conversion



Empowering 200+ e-commerce merchants
How do Performance Max experiences work?
Experiences are a feature already known on the advertising platform. The principle: you split a portion of traffic into two groups. The 'control' group keeps the current configuration, while the 'test' group receives the modification you want to evaluate. You will then know which performs better, backed by figures.
The big novelty is finally being able to launch this type of test on Performance Max. This was not possible before, and it changes the game for merchants who doubted the value of the format.
You can, for example, pit a bidding strategy targeting cost per lead against another targeting return on ad spend. Two different logics, one measured winner.
There are three types of experiences, presented in the following sections. Each answers a specific question about the real value of your campaigns.
Can you compare a Shopping campaign to Performance Max?
Yes, and this is the first type of experience available. You pit a Performance Max campaign against another type of campaign: Shopping, Search, or Display. The test is done on a 50/50 traffic split.
Concretely, if you are already running a Shopping campaign, half the traffic continues to go to Shopping and the other half to Performance Max. After a few weeks, you know which of the two mechanisms converts better for the same budget.
This comparison is valuable for a merchant attached to their existing campaigns. It avoids migrating blindly to Performance Max just because Google recommends it.
You decide based on real data, not intuition. And if Shopping wins, you keep your current structure with full knowledge. To go deeper, see how to structure your category pages for this type of campaign.
What is the purpose of the experience on the final URL extension?
Second type of experiment: the final URL extension test. This feature lets Google choose, on your behalf, the destination page it deems most relevant for a given query.
In the test, half of the traffic goes through this automatic optimization while the other half stays on the URLs you selected yourself. The goal: to see if your manual choice performs better than the algorithm's.
This is a legitimate question for any e-commerce business. A well-structured category page can rival the product page that Google would prefer to promote. The reverse is also true.
The test gives you the answer instead of leaving the option enabled by default. If your manual URLs win, you disable the extension and regain control over the arrival experience. Combined with conversion rate optimization work, this setting has a significant impact on your return on investment.
How to measure the real contribution of Performance Max?
The third type of experiment is called Uplift. It is the most interesting because it measures the incremental gain brought by Performance Max when running alongside your other campaigns.
The principle remains a 50/50 traffic split. The control group receives only your usual campaigns. The test group receives these same campaigns plus a Performance Max campaign or comparable campaigns.
The difference between the two groups is the net gain attributable to Performance Max. You no longer guess; you measure. This is precisely what was missing to calmly decide on this format.
Beware of one point: this test requires strict tracking discipline. Without a reliable analytics tool, you risk comparing two dissimilar periods. Therefore, cross-reference these results with a suivi analytique propre.
What are comparable campaigns in an experiment?
For the Uplift test to make sense, Google groups under the term of comparable campaigns those that pursue similar objectives. These campaigns are modifiable, giving you a certain degree of control over the test scope.
Three criteria define a comparable campaign: sharing the same domain, targeting one or more overlapping conversion objectives, or targeting overlapping geographic areas.
Concrete example: your Performance Max campaign targets winter jackets. A Search campaign on jackets and a Demand Gen campaign on a winter theme will be considered comparable.
You thus build a coherent test group that speaks to the same audience. Without this coherence, the result would be biased and you would not know what to do with it.
Why exclude audiences with your data?
Second major innovation: data exclusions. For years, Google allowed excluding keywords and locations, but not customer match lists or remarketing lists.
Now, a campaign setting named 'Your data exclusions' integrates customer match and remarketing audiences. You can therefore prevent certain audiences from seeing your ads.
For a merchant, the most common use is to avoid paying for people who are already customers, or for segments that never buy. On acquisition campaigns, this filter protects your budget.
However, the need to exclude varies greatly from one advertiser to another. What works for one store may harm another. Test before generalizing.
When should you avoid excluding an audience?
Audience exclusion is not a universal reflex. In practice, the right setting depends on your model, your margin, and your sales cycle.
A supplement brand may want to exclude recent buyers to avoid cannibalizing its follow-ups. Conversely, a gift shop often targets the same people from one season to the next.
Another nuance: on Performance Max, excluding an audience reduces the algorithm's hunting ground. Too many exclusions can slow down the learning phase and increase the cost per acquisition.
The right approach: exclude gradually, one audience at a time, and observe the real effect on your sales. To frame this logic on social networks, draw inspiration from our advice on a e-commerce Facebook Ads strategy.
How to spot product overlap in your account?
Third novelty: product overlap. It identifies products appearing in multiple Shopping campaigns in your account. This isn't specific to Performance Max, but it's very useful for cleaning up a structure that has grown.
To access it, open the 'Campaigns' section on the left, then click on 'Products'. You get the full list of your products with associated data.
Click on a specific product: its attributes are displayed, accompanied by a dropdown menu listing the campaigns containing it. You immediately see where it is shown and at what price.
You can then review results by campaign and exclude underperforming items. It's the same logic as applying negative keywords to trigger the right ad.
How to arbitrate between your overlapping campaigns?
Spotting an overlap is good. Deciding what to do about it is better. A product shown in three different campaigns can dilute your budget instead of concentrating it where it generates returns.
The method: compare the performance of the same product, campaign by campaign. Keep the one showing the best return, cut or reduce the others.
Be careful not to delete everything at once. Some campaigns serve to cover search queries that Performance Max doesn't reach. The goal is to complement, not replace.
This discipline aligns with that of a e-commerce strategy designed for a small brand : few campaigns, well-maintained, rather than a difficult-to-read pile.
How does Qstomy help maximize Performance Max ROI?
A Performance Max campaign improves traffic. But the traffic must convert into revenue. This is where Qstomy comes in, an AI agent designed for Shopify stores.
Qstomy guides the buyer toward a decision: product recommendations, upselling, cross-selling, cart management, order tracking, and after-sales service. In other words, the entire journey that turns an expensive ad click into an order.
Used by over 100 merchants, the agent acts where advertising stops. You pay to attract attention, then Qstomy works on conversion, average order value, and loyalty, with clear answers on orders, tracking, and return policies.
Result: your Performance Max tests are easier to justify, as every euro invested in acquisition finds a continuation in sales. To complement, read our approach to advertising strategy after iOS updates.
Quelle checklist avant de tester Performance Max en 2026 ?
Before launching your experiments, review these few points. They prevent you from drawing false conclusions from poorly prepared tests.
Trafic suffisant : a test needs volume to be reliable.
Suivi propre : check your conversion goals before comparing.
One variable at a time : do not test bids and audiences simultaneously.
Realistic duration : allow time for the learning phase to stabilize.
Written decision : note what you keep and what you cut.
In short
Experiments compare Performance Max to Shopping, Search, or Display, or measure its incremental gain. Audience exclusions filter your customer match and remarketing audiences. Product overlap identifies duplicates in your Shopping campaigns. Three control levers, to be activated methodically.
FAQ
Should you test everything at once? No. An isolated test per subject yields actionable results.
Are audience exclusions always useful? No, it depends on your model and sales cycle.
Is product overlap specific to Performance Max? No, it concerns all your Shopping campaigns.
To go further, compare the platforms with our guide on the choosing an e-commerce platform.

Enzo
October 28, 2026


