Performance Max or Standard Shopping: decide by product group

You choose per product group

The question "PMax or Standard Shopping" is asked at the wrong level when it is asked for the whole shop. For a shop with working tracking and around 60 conversions a month or more, I would start with Performance Max as a rule and only use Standard Shopping where you can name a specific action that is not possible in PMax. "More control" is not such an action, "capping the CPC at €0.15" is.

Two things support this order. Among Tinuiti's advertisers, PMax accounted for 67% of Shopping spend and 68% of revenue in the first quarter of 2026, with a slightly better ROAS than Standard Shopping for brands using both. That is not a figure for the DACH region (Germany, Austria and Switzerland), nor evidence that this revenue was additional. And the software vendor smec, which itself sells a hybrid product, writes that most retailers probably do not need a hybrid. The statement runs against its own product, which is why I give it more weight.

Two hard differences remain

The first concerns placements. Standard Shopping runs on the Shopping tab, Search, Google Images and optionally search partners, while PMax additionally reaches YouTube, Display, Gmail and Maps, among others. That can mean reach, or it can mean spend you are unable to evaluate. What matters to the retailer is whether they want to pay for these placements in this product group.

The second concerns the bid. Only Standard Shopping offers Manual CPC and Maximise clicks. PMax works exclusively with automated bidding for conversion value or conversions. You cannot set a fixed cost per click there.

Two older arguments, on the other hand, no longer hold. The PMax search terms report shows the search terms that triggered ads, and negative keywords apply at account or campaign level. Query visibility alone is therefore no reason for Standard Shopping, and the details are under search terms in Performance Max. And an automatic PMax priority no longer exists, more on that below.

The decision table

The table is my assessment. Work through it from top to bottom for each product group. If a group does not clearly move to the right in any row, it stays in PMax.

The second row contradicts the common advice. Margin spread alone does not argue for Standard Shopping, because separate PMax campaigns also have their own targets. The implementation is described in target ROAS per margin group. Only the combination of spread and thin data changes the answer. That is also why blanket rules such as "best sellers in Standard Shopping" or "low-margin items in Standard Shopping" help little: a low-margin best seller has enough data for its own target, while a high-margin niche item may not.

Worked example: margin first, then data

Suppose a shop makes 200 conversions in 30 days, spread across three margin groups. The break-even ROAS is 1 divided by the contribution margin rate. All figures are invented.

A and B do not belong under one target. With a blended target of 3.5 that both hit exactly, A stays in profit, while B brings €350 of revenue and €87.50 of contribution margin per €100 of ad spend, which is a loss of €12.50. The margin tells you that the groups need separating. It does not yet tell you anything about the campaign type.

That only comes with the data check for C. For Target ROAS in Search and Shopping, Google requires at least 15 conversions in the past 30 days at the conversion tracking level, not per campaign. Formally, C could therefore get a target. I still think it makes little sense to base a bid on a segment with ten conversions, because three orders more or fewer already move the ROAS visibly.

For C, the hard cap is a better fit. The formula: break-even CPC = order value × contribution margin rate × conversion rate. With an assumed order value of €80, a 12.5% margin and a 2% conversion rate, that comes to €0.20. At that price nothing is left after ad spend, so the cap belongs below it. Estimate the conversion rate over a longer period or from the parent category, not from the ten conversions. If the group gets no meaningful clicks at this price, you should not advertise these products at this margin.

Every split has its price. According to Google, performance generally improves with fewer, larger campaigns, which fits the data logic. On top of that, smec estimates the extra workload of a manually run hybrid at 30 to 50%.

When Standard Shopping clearly wins

The clearest case is the one from the example: a thin segment whose profitability depends on a fixed click price. No target can do that, because a target ROAS aims at an average and does not limit any individual price.

It is just as clear when you deliberately want to buy only Shopping reach for a product group, for example because you cannot evaluate YouTube spend for these items or do not want to pay for it. Since Google introduced AI Max for Shopping, this is also possible with more automation: product controls and bidding options remain, and without Final URL expansion it stays with Shopping ads. For performance across channels, Google continues to point to PMax.

The third case is short-term clearance of stock with a weak history. A manual bid needs no learning period and does not weight the past. The catch: someone has to maintain the bid and end the campaign once the stock is gone.

What the hybrid looks like, and who gets priority

Running both types in parallel is possible. In October 2024 Google removed the automatic priority of PMax, and the campaign with the highest Ad Rank serves the auction. If you still read about PMax taking priority, you are looking at outdated information. According to Google's help pages, this is meant to stop campaigns in the same account from driving up each other's bids. There is no switch for priority: Google recommends steering spend through budgets and targets. There is one exception with Search, because an exact match keyword that exactly matches the search query is preferred over PMax.

My rule: every product in exactly one Shopping or PMax campaign. If you run the same product twice, you leave the choice of bid to Ad Rank and can hardly read the results of the two campaigns separately any more. For the example above, this gives a lean structure:

• PMax A with a target above 2.5 for the high-margin group
• PMax B with a target above 4.0 for the middle group
• Standard Shopping C with Manual CPC and a cap below €0.20, only for the thin group
• Assignment via a label in the feed, so that no product runs twice

When you create each Standard campaign, write down the action it exists for. If you have never carried out that action after three months, the group goes back into PMax. The deadline is my suggestion.

Tracking and margin come first

Without correct conversion tracking, or if you do not know the margins per product, break-even ROAS and CPC caps are false precision, and tracking comes before any question of structure. It also remains open whether a hybrid really brings additional profit. I know of no robust incrementality study on this, and the reported ROAS does not answer the question. If you want to know for your account, you have to test PMax against Shopping properly. Feature details as of November 2026.