Performance Max Reporting: Channels, Search Terms & Profit

What the reporting now shows

Advice written a couple of years ago says Performance Max shows neither channels nor queries. That is no longer accurate. Google publishes channel performance reporting for Performance Max and a search terms report. Both are worth reading before you conclude anything about a campaign.

Channel performance sits under Campaigns > Insights and reports > Channel performance in Google Ads. It reports across Google Search, Display, YouTube, Discover, Maps, Gmail and Search partners. Shopping is not a separate line: the segment covering ads that use your product data is broader than Shopping ads alone, so reading it as a Shopping figure overstates what it says. The placement report is a separate view, and it carries impressions only.

The useful question is no longer "what can I see". It is "what does each report answer, and what does it not settle". Reported distribution is not causal proof. Impressions on a placement do not establish what that placement cost you or earned you. A share of new customers is not a measurement of incremental sales.

Report, question, limitation

A reporting workflow, in order

1. Confirm what the campaign is optimising for before reading any performance number: the conversion goals attached to it, and whether they are the ones the business is judged on.
2. Fix the date range and the attribution settings, and keep them fixed for every report in the same review.
3. Read channel performance for distribution. Note where activity concentrated, and treat it as a question to investigate rather than an allocation verdict.
4. Read the search terms report for matched demand, remembering that terms below Google's thresholds are not listed.
5. Read asset group and listing group reporting to see which products and which structures carried the spend.
6. Export the orders for the same period from your own systems and reconcile them to the platform's reported conversions.
7. Apply product-level cost to those orders and read contribution, not revenue.
8. Write down the decision, the evidence behind it and the date, before making any change.

Validating conversion goals and values

Plenty of Performance Max arguments turn out to be arguments about the conversion action. Before comparing anything, check which goals the campaign counts, whether any of them double count the same order, and whether the value passed is the order value you recognise.

Currency is a question of accuracy rather than uniformity. The currency sent with a conversion should represent what the customer was actually charged; Google can convert a reported currency into the account currency, so a currency that differs from the account is not automatically wrong. What breaks reporting is a value or a currency that misstates the order.

A campaign optimising toward an inflated or duplicated value will do exactly what it was told to do. That is a measurement fault rather than an algorithm fault, and channel reporting is unlikely to surface it on its own.

Consistent dates and attribution windows

Comparisons break most often because two numbers were produced under different rules. By default, Google Ads reports a conversion against the date of the ad interaction that led to it, while the by-conversion-time columns report it against the date the conversion happened. Your ecommerce platform reports orders against the date of the order, and a conversion window keeps counting after the period ends.

Pick one date basis, state the attribution window, hold both constant across the period being compared, and note them next to the figures. If the window changes, the series before and after it are not comparable.

Reconciling to product and order contribution

The platform reports revenue it is credited with. Your P&L reports what remained. Reconciliation is the step that connects them: match the orders, identify the ones the campaign is credited with, then apply cost of goods, shipping, payment fees and returns at product level.

Expect the two sides not to agree exactly. Attribution rules, modelled conversions and fractional credit mean reported conversions do not map one to one onto individual orders, so treat reconciliation as narrowing the gap rather than closing it. Record the size of the unresolved difference and the reason you believe it exists, then judge trends against that stated tolerance.

That is where a reported return figure turns into a contribution figure you can act on. Our explanation of POAS covers the arithmetic, and Performance Max versus standard Shopping covers the structural choice this reporting usually informs.

What needs a test rather than a report

Reports describe distribution rather than establish incrementality. If the question is whether spend caused sales that would not otherwise have happened, the answer comes from a designed test with a control: a holdout, a geographic split, or a deliberate change measured against a comparable group.

Design the test around the decision. The period, the size of the change and the threshold that would count as a result all follow from your order volume and your purchase cycle. There is no fixed 28-day, 25% recipe that makes a test valid, and separating brand traffic into its own campaign is a structural change rather than an incrementality proof.

Brand overlap is a real question and worth answering properly: how we approach commercial measurement sets out the evidence we ask for before acting on it.

The decision log

Keep a short log of every material change: the date, what changed, what evidence prompted it, and what you expected to happen. It costs a line of text and it makes the next review far easier to interpret, because it records what you did alongside what the market did. It is a record, not a causal test.

Without it, a campaign carrying six months of undocumented edits is much harder to read: a trend has several possible causes and no record of which applied.

FAQs

Related reading

• Performance Max visibility hub: the wider guide this article sits under.
• The hidden cost of Performance Max opacity
• Performance Max management