The choice between Performance Max and Standard Shopping is usually framed as a binary, and that framing is why so many accounts get it wrong. They are not competing versions of the same campaign type. Standard Shopping gives you control over which products get spend and what you can exclude; Performance Max gives you reach across surfaces you cannot buy any other way, in exchange for visibility and control. The account that wins is the one that matches the choice to its catalogue, its data volume and its margin structure. For the Foto Ruano campaign that returned x49 ROAS, the answer was Standard Shopping segmented by margin, which is the opposite of what most agencies would have deployed.
Quick reference — choosing between them:
- PMax takes priority over Standard Shopping for the same product in the same account, regardless of bid
- Standard Shopping wins when margin varies widely across the catalogue
- PMax wins when the catalogue is large, homogeneous and conversion data is plentiful
- Brand exclusions are mandatory in PMax or it will absorb brand traffic and take credit for it
- A valid test needs a product-set or geo split, not a before-and-after comparison
- Neither is a fix for a broken feed or unreliable conversion tracking
What actually separates the two campaign types?
Standard Shopping is a product-feed campaign that runs on the Shopping surface and Search partners. You control which products are in which ad group, you can bid differently by product group, you can see and add negative keywords, and search terms reporting tells you something useful about what triggered spend.
Performance Max is a cross-channel campaign that uses your feed plus a set of creative assets to serve across Search, Shopping, Display, YouTube, Discover, Gmail and Maps from a single budget. Google decides the channel mix, the placements and much of the targeting. You supply the feed, the assets, audience signals and conversion goals.
The critical mechanical detail that catches accounts out: when a PMax campaign and a Standard Shopping campaign in the same account are eligible for the same product, PMax serves. This serving priority is documented in Google’s Performance Max help documentation, and bid level does not override it. So “running both” on overlapping inventory does not create a fair contest; it quietly starves the Standard Shopping campaign and produces data that looks like PMax winning when the auction was never competitive.
When does Standard Shopping outperform?
Standard Shopping earns its place whenever the account needs the spend distributed deliberately rather than optimally-on-average.
Wide margin variation across the catalogue. This is the strongest case, and it was the deciding factor on Foto Ruano. When a catalogue contains items at 60% margin and items at 12% margin, a single blended ROAS target will overspend on the cheap-to-convert low-margin items and underspend on the profitable ones. Splitting products into margin bands with different ROAS targets fixes an economics problem that PMax has no mechanism to express. Our Google Shopping guide for Australian e-commerce covers how to build that segmentation.
Small or seasonal catalogues. With a few dozen SKUs, PMax’s machine learning has little to optimise across and the loss of control buys you nothing.
Thin conversion data. PMax needs conversion volume to learn. An account generating a handful of conversions a month gives it nothing to work with, and the campaign spends its way through a learning phase that never completes.
A hard performance window. When a campaign has a fixed spend period and a target it must hit, as in a Black Friday or EOFY push, you cannot afford to spend the first two weeks letting an algorithm learn its way through a funnel. Standard Shopping is predictable in a way PMax is not.
Diagnostic need. When something is wrong and you need to see what queries and products are consuming budget, Standard Shopping shows you considerably more.
When does Performance Max genuinely win?
PMax is not a downgrade, and treating it as one costs accounts real money.
Large, homogeneous catalogues. Thousands of SKUs at broadly similar margins is where PMax shines. There is genuine value in an algorithm allocating across inventory at a scale no human will manage manually.
Strong, reliable conversion data with values attached. PMax rewards accounts feeding it accurate revenue values. If your conversion tracking is solid and passing transaction values, PMax has the signal it needs. Google’s Smart Bidding documentation sets out the conversion volume automated strategies need before they optimise reliably.
Demand beyond Search. The channels PMax reaches, YouTube, Discover, Gmail, Display, are genuinely additive for consumer products with visual appeal and a discovery-driven purchase. You cannot replicate that mix with Standard Shopping at all.
Remarketing at scale. PMax handles returning-visitor demand well when audience signals are properly configured.
The honest summary: PMax is strong at finding incremental demand across surfaces, and weak at respecting constraints you cannot express to it.
Why brand exclusions are not optional
Left unconfigured, a PMax campaign will bid on your brand terms. Brand traffic converts at a high rate and a low cost, so PMax absorbs it, reports excellent blended performance, and the reported ROAS reflects demand that already existed rather than demand the campaign created.
This is the single most common reason a PMax campaign appears to outperform in a head-to-head test. The campaign is not finding new customers; it is claiming credit for people already searching your name.
Apply brand exclusion lists to every PMax campaign before you evaluate performance, and judge PMax on non-brand results only. It is the same principle behind reporting non-brand ROAS separately, which is one of the mistakes we see most often across audited accounts: blending brand and non-brand produces a flattering number that hides what is actually happening.
How to run a test that produces a real answer
Most PMax-versus-Shopping tests are worthless because they compare the wrong things. A before-and-after comparison confounds the campaign type with seasonality, price changes, stock levels and competitor activity. Running both simultaneously on the same products is invalidated by PMax’s serving priority.
Two designs work.
Product-set split. Divide the catalogue into two comparable halves, matched on margin band, price range, category mix and historical conversion volume. Run PMax on one half and Standard Shopping on the other. Because the campaigns cover different inventory, PMax’s priority never triggers. Compare non-brand ROAS and total profit, not revenue.
Geographic split. Run PMax in one set of states or metro areas and Standard Shopping in another, matched on historical volume. This keeps the full catalogue in both arms, which is useful when your catalogue is too small or too varied to split cleanly. It requires enough volume in each region to reach significance.
Whichever you choose, hold it for at least four to six weeks. PMax needs time to exit its learning phase, and a test that ends at two weeks measures the learning phase rather than the steady state.
Measure profit, not ROAS. If margin varies across your catalogue, a campaign can win on ROAS while losing on contribution profit by shifting spend toward low-margin, easy-to-convert products. Work out the return you actually need with our break-even ROAS calculator before deciding which arm won.
The hybrid that usually works
For most established e-commerce accounts the answer is not one or the other, but a deliberate split of inventory.
Put your high-margin, strategically important, or tightly-managed products in Standard Shopping where you control bids and exclusions. Put the long tail, the homogeneous bulk of the catalogue, into PMax where automated allocation genuinely helps. Because the two campaigns cover different product sets, the priority rule never puts them in conflict.
This gives you algorithmic reach across the catalogue you cannot manage by hand, and human control over the products where margin, stock or strategy demands it.
Frequently asked questions
Does Performance Max override Standard Shopping in the same account? Yes. When both campaigns are eligible to serve for the same product, PMax takes priority regardless of bid. This is why running both on overlapping inventory does not produce a valid comparison, and why a hybrid setup must split products between the campaigns rather than duplicating them.
Can I still use Standard Shopping in 2026? Yes. Standard Shopping remains fully available and is the better choice for catalogues with wide margin variation, small SKU counts, limited conversion data, or campaigns with hard performance windows. Reports of its removal have been circulating for years and remain inaccurate.
How long should a PMax versus Standard Shopping test run? Four to six weeks minimum. PMax needs time to exit its learning phase, and shorter tests measure that phase rather than steady-state performance. Use a product-set or geographic split rather than a before-and-after comparison, which confounds campaign type with seasonality and stock changes.
Why does PMax always look better in my reporting? Usually because it is absorbing brand search traffic. Brand terms convert cheaply and at high rates, so a PMax campaign without brand exclusions reports strong blended performance built on demand that already existed. Apply brand exclusions and compare non-brand results before drawing conclusions.
Is Performance Max good for lead generation? Much less reliably than for e-commerce. Without a product feed and with lead quality harder to signal back to the algorithm, PMax frequently optimises toward high volumes of low-quality leads. Our detailed assessment of Performance Max covers when to avoid it for lead gen entirely.
Which produces better ROAS overall? There is no universal answer, and any source claiming one is not accounting for catalogue structure. PMax tends to win on large homogeneous catalogues with strong conversion data; Standard Shopping tends to win where margin varies widely or control matters. Measure contribution profit rather than ROAS, because a campaign can improve ROAS while reducing profit.
The productive question is not which campaign type is better, but which parts of your catalogue need control and which benefit from reach. Accounts that answer that honestly usually end up running both, on deliberately separated inventory, and outperform the accounts that picked a side.
If you are deciding how to split your catalogue or your PMax numbers look better than your bank account does, get in touch and we will work through the structure with you.