PMax vs Standard Shopping Attribution: Which Campaign Really Gets the Credit?

Picture an account running Performance Max and Standard Shopping side by side. PMax reports 600% ROAS. Standard Shopping reports 280%. Total revenue has not changed. So which campaign is really producing the sales?

It is one of the most common questions in PPC forums at the moment, and the honest answer is that it is harder to tell than Google’s reports make it look. Here is why, and what you can do about it.

Why Attribution Gets Messy With More Than One Campaign

When someone clicks several ads before buying, Google Ads uses an attribution model to decide which click gets credit. The default is data-driven attribution, which shares credit across the touchpoints in the path based on how much each one is predicted to have contributed.

With PMax and Standard Shopping running together, the same shopper can interact with both before converting. Under a data-driven model, both campaigns may get part of the credit for that one sale. That is not wrong (both touchpoints played a part), but it does mean each campaign’s conversion column is not a clean, isolated measure of its own performance.

PMax adds a specific complication. It covers both upper funnel touchpoints (Display, YouTube impressions, Discover) and lower funnel ones (Shopping, Search). If a shopper sees a PMax Display ad early on, leaves, later clicks a Standard Shopping ad and buys, both campaigns share the credit. Yet Standard Shopping was the final click that sealed the decision, and PMax’s share comes from an early impression that may or may not have mattered.

Why PMax ROAS Can Look Great While Sales Stay Flat

A few forces combine to produce this.

It captures your brand searches. Without brand exclusions, PMax bids on searches for your store name. Someone typing “yourstorename.com” already intends to buy. PMax takes the click, records the sale and claims full credit for revenue that was coming anyway. Reported ROAS goes up without any genuine extra revenue from advertising.

It cannibalises retargeting. PMax spreads budget across every inventory type, including retargeting on Display and YouTube. Someone who abandoned a basket yesterday was very likely to buy whether or not they saw a retargeting ad. PMax can pick up these high converting users, report excellent ROAS on them, and spend budget that could have found new customers.

The attribution model hands it credit. As explained above, data-driven attribution can give PMax a sizeable share of conversions that other campaigns or organic channels would have delivered regardless.

Together, these effects can push PMax’s reported ROAS well above its true incremental ROAS. And because PMax does not show placement-level data, working out how much of its reported revenue is genuinely incremental is difficult.

Getting a More Accurate Picture

You cannot perfectly measure PMax’s incremental ROAS without a controlled experiment. You can, however, make the reported numbers far more meaningful.

Add brand exclusions. This is the most important step by far. If PMax is not excluding brand terms, its ROAS is inflated by brand traffic. After you add exclusions and let the campaign run for two to three weeks, reported ROAS usually drops, but the new figure is a much more honest view of what PMax does with non-branded budget.

Judge total account revenue, not campaign ROAS. The useful question is not “what is PMax’s ROAS?”. It is “when PMax runs at this budget, does total store revenue rise compared with when it was off, or running at a lower budget?”. That means looking at revenue over time and at account level, not at campaign metrics.

Run a holdout experiment. Google Ads campaign experiments let you draft your current structure and split traffic 50/50, so one group sees PMax and the other does not. The revenue difference between the groups is your best estimate of PMax’s real incremental impact. You need enough traffic for statistical significance and a few weeks of runtime, but it is the only way to get a properly clean answer.

Check GA4 alongside Google Ads. GA4 applies its own attribution model (data-driven by default, but worth confirming). Comparing how GA4 and Google Ads assign credit shows you where the two models disagree and gives a fuller picture of the conversion path.

When PMax ROAS Is High but Revenue Is Flat

Work through these questions:

Is PMax spending heavily on brand terms? Look at the search term categories in the PMax insights report. If brand terms dominate, add brand exclusions straight away.

Has Standard Shopping spend fallen since PMax launched? PMax takes priority over Standard Shopping when both are eligible for the same product. If Shopping spend dropped noticeably when PMax went live, PMax may be cannibalising it rather than adding new reach.

Is the conversion window long? With a 30 or 90 day window, PMax may be claiming purchases that began well before anyone saw its ad. Comparing 7 day and 30 day conversion data shows how much reported ROAS comes from long window attribution.

Has PMax been live long enough to finish learning? For the first four to six weeks it is still working out how to split budget across inventory types, so results in that window are not representative.

Comparing PMax and Standard Shopping in Looker Studio

A side by side view in Looker Studio makes it easier to compare campaigns without flicking between Google Ads screens.

Connect the data:

Add your Google Ads account as a data source. The native Google Ads connector exposes campaign-level data including campaign name, campaign type, impressions, clicks, cost, conversions and conversion value.

Build the comparison table:

Create a table with Campaign as the dimension and these metrics as columns:

• Cost
• Conversions
• Conversion Value
• Cost per Conversion (calculated field: Cost / NULLIF(Conversions, 0))
• ROAS (calculated field: Conversion Value / NULLIF(Cost, 0))
• Click Share (if the connector provides it)

Filter the table to your Shopping-related campaigns (PMax and Standard Shopping), for example Campaign Name contains “Shopping”, or whatever naming convention you use.

Add a time comparison:

Use a date range comparison (this month against last month, or the period since PMax launched against the same length of time before) to see how metrics have moved. A table showing PMax ROAS this period against last period, next to total account revenue for the same periods, exposes the gap between what campaigns report and what the business actually earns.

The most revealing chart:

A daily time series of total account conversion value plotted against PMax’s daily spend is often the single most telling view. If PMax spend climbs while total conversion value stays flat, that strongly suggests PMax is not adding revenue. It is moving credit away from other channels.

The Structural Choice: PMax Plus Standard Shopping, or Just One?

When the same products are eligible in both PMax and Standard Shopping, PMax wins the auction by default. Standard Shopping only shows on queries PMax did not bid on, which is essentially leftover inventory.

In most accounts, running both without careful product exclusions leaves PMax dominant and Standard Shopping as a minor extra. Asking which one performs better then has no clear answer, because they are not competing on equal terms.

A cleaner setup:

• Run PMax for the full catalogue with brand exclusions
• If you want Standard Shopping coverage, exclude specific product groups from PMax and keep those products in Standard Shopping only
• You now have a genuine comparison: different products, different campaigns, no overlap

It takes more effort to set up, but you get real comparative data instead of the blurred view you get when two campaigns fight over the same products.

The alternative is to run PMax alone for 90 days, collect its performance data, then decide whether adding Standard Shopping is worth the extra complexity for your account. Plenty of stores running PMax on its own, with good signals, brand exclusions and clean conversion tracking, find it performs well without the overhead of managing two campaign types at once.