Performance Max revenue rises while Search revenue falls by almost the same amount. Has the business gained new demand, or has demand moved between reporting columns? The campaign dashboard starts the investigation; it cannot answer the question alone.
Performance Max, or PMax, provides access to multiple Google channels and can complement keyword-based Search campaigns. That describes its role and capabilities; suitability still depends on the business’s goals, measurement and assets. [1]
1. Define each campaign’s job
You might use Search to cover explicit demand for a service and test PMax across additional inventory. Record the intended outcome, assigned value, products, markets and constraints the client needs to preserve.
Check that the campaign types use comparable definitions. A completed order in one and an initial form submission in another are not interchangeable revenue. Fix currency or value inconsistencies before discussing budget.
2. Separate attributed results from business outcomes
Attribution connects conversions to advertising interactions. First ask what the platform credits to each campaign. Then ask what changed in total business performance and the product/customer mix. Finally, ask whether a test can estimate the additional sales caused by advertising.
Review brand search coverage and repeat demand within available data, documenting its limits. Higher ROAS does not automatically mean better new-customer acquisition. Nor should you assume an exclusion or separation setting behaves identically in every account.
Practical example: one column rises—what happened to the total?
These are hypothetical figures, not client results.
| Hypothetical period | Revenue attributed to Search | Revenue attributed to PMax | Total revenue in the business’s records |
|---|---|---|---|
| Before changing allocation | 8,000 USD | 4,000 USD | 13,000 USD |
| After changing allocation | 6,000 USD | 6,000 USD | 13,200 USD |
PMax’s attributed revenue increased by 2,000 USD; the business total increased by just 200 USD. This alone does not prove demand shifted between campaigns. It does justify avoiding the assumption that one column’s growth equals business growth.
Add spend, margins, order quality and conversion lag, then review seasonality and the offer. A changing mix can alter how an average should be interpreted even when overall revenue barely moves.
3. Choose a test for a specific question
Define the change, comparison unit, constants and guardrail metrics. Google offers different experiment types, but support varies by campaign type and configuration. Check eligibility; do not assume a custom Search experiment automatically tests PMax. [2]
If you cannot isolate the change, present a descriptive comparison with a clear record of other factors. It can support an operational decision, provided its limitations are stated rather than presented as causal evidence of incremental impact.
4. Present budget options with conditions
Options include continuing, testing a limited increase, fixing measurement or reallocating gradually. For each, state what you will monitor, when you will review and when you would reverse course.
The useful client deliverable is an explained decision: what is known, what will be tested and how order value and quality will be protected. Picking a “winning campaign” from two ratios makes a shorter report, but does not settle the budget question.
Official sources
Source documentation reviewed: 16 October 2026. Features and eligibility may change; check the linked documentation before implementation.
