Microsoft Ads: New PMax experiment types

Microsoft is significantly expanding testing options for Performance Max campaigns in Microsoft Advertising. Advertisers receive two new experiment types to validate changes to automated campaigns in a controlled way and measure incremental impact without jeopardizing live campaign performance. For teams running Performance Max alongside Google Ads or as a core acquisition channel, Microsoft is closing an important platform gap.

Performance Max is one of the most heavily automated campaign formats in paid search. Algorithms manage bids, audiences, and placements across multiple inventory sources. That automation makes it hard to evaluate individual adjustments cleanly. Without dedicated experiments, many optimizations remain guesswork: Is revenue rising because of a new asset group, a changed target CPA strategy, or simply seasonal demand? Microsoft now addresses this with structured test frameworks directly in the ads interface.

Two new experiment types at a glance

The update introduces two clearly separated formats that answer different questions. Uplift experiments measure the incremental value of a Performance Max campaign against a control group. Upgrade experiments compare an existing campaign with a revised Performance Max version before changes are fully rolled out. Both approaches are available for eligible accounts under Campaigns > Experiments.

• Uplift experiments: They show the additional effect Performance Max delivers versus a defined control setup, for example when justifying budget or evaluating channel value.
• Upgrade experiments: They enable a direct comparison between a live campaign and an optimized variant, for example after asset updates, goal changes, or feed adjustments.

Why this matters for advertisers

Until now, experiments in Microsoft Ads were limited to search campaigns. Teams relying on Performance Max often had to test changes directly in live campaigns or rely on external analysis. That increases risk and makes causal statements harder. By extending testing to Performance Max, advertisers gain a more controlled way to validate campaign changes, optimize performance, and make budget decisions based on data before committing larger spend.

Incrementality is a recurring topic in automated formats. Platforms report strong conversions, yet without clean control groups it remains unclear which share truly comes from paid media. Uplift tests help make that difference visible and support internal discussions on channel value or budget shifts with more reliable data. For agencies and in-house teams that must defend Performance Max in quarterly reviews or budget planning, that is a meaningful step beyond simple before-and-after comparisons in live accounts.

Renaming existing search experiments

Alongside the expansion, Microsoft renamed its previous experiment offering for search campaigns. It is now called Search optimization experiments, clearly separated from the new Performance Max testing capabilities. The distinction reflects a broader platform approach: Microsoft wants to provide advertisers with more sophisticated optimization tools across automated campaign formats and structure the experiment landscape more clearly.

Uplift experiments in detail

Uplift experiments target teams that want to quantify the marginal contribution of Performance Max. Typical use cases include budget approvals, channel comparisons with other paid media investments, or the question of whether a campaign delivers incremental revenue at all. The control group forms the reference frame; differences in conversions, revenue, or cost per acquisition can be evaluated systematically. For account managers, that means less blind flying in automated setups.

Upgrade experiments in detail

Upgrade experiments suit scenarios where a live Performance Max campaign already exists and an improved variant should be tested. Instead of rebuilding a productive campaign directly, the new version can run in parallel and be measured against the existing one. That reduces performance risk during larger changes, such as after feed optimizations, new creatives, revised target CPA values, or conversion goal adjustments. Only when the test variant performs well does full rollout follow.

Practical recommendations for Microsoft Ads teams

Teams that want to use the new features should first check whether the account is eligible for experiments. Next, it helps to define clear hypotheses: Is the focus on incrementality or on comparing two campaign states? Upgrade tests suit operational optimizations; uplift tests suit strategic budget and channel questions. Run times and budget shares should be large enough for algorithms and conversion data to deliver meaningful signals. In practice, it also helps to change only one core variable per test so results remain clearly interpretable.

Microsoft positions the expansion within a broader optimization offering for automated campaigns. While Google Ads has long provided extensive test and reporting options for Performance Max, advertisers on the Microsoft side are now catching up. For multichannel setups, that means more comparability when evaluating upgrades and a more consistent experiment culture across platforms.

• Define a measurable hypothesis and primary KPI before launch.
• Use upgrade experiments for asset, feed, or bid changes.
• Use uplift experiments for questions about incremental campaign impact.
• Reconcile results with CRM or analytics data, not only platform metrics.
• Do not mix Search optimization experiments with Performance Max tests.

The update was first spotted in Microsoft Ads help documentation and confirms the trend that major ad platforms are equipping automated formats with controllable test environments. For Performance Max owners, that means more operational safety when upgrading campaigns and more reliable measurement of true campaign impact in Microsoft Advertising.