The short answer: there is no universal winner between Performance Max and Standard Shopping. Choose Standard Shopping when you need product-level control, query visibility, and a clean baseline you can reason about. Choose Performance Max when your conversion values are trustworthy, your feed is healthy, you can supply real creative beyond product images, and you are willing to trade some control for reach across Google's inventory. If you cannot yet say which is better for your catalogue, do not decide by comparing the two campaigns side by side in the interface. Run a structured campaign experiment, and judge the result on contribution after returns and variable costs, not on platform-reported ROAS.
This guide is for DTC founders, heads of growth, and paid search leads deciding how to allocate Shopping budget. It compares the two campaign types on the same eight criteria, then walks through four gates that should be passed before either option earns more budget. If you already run Performance Max and need to decide whether to keep it, our Performance Max audit framework covers that separate job.
What each campaign type actually is
Standard Shopping is a campaign type that serves product ads from your Merchant Center feed on Shopping placements. You structure it through product groups, you set bids or a bid strategy, and you can review which search terms triggered your ads. It is narrower by design: its job is Shopping ads, and most of its controls live at the product and query level.
Performance Max is a goal-based campaign type that can serve across Google's inventory, including Search, Shopping, YouTube, Display, Discover, Gmail, and Maps, from one campaign (Google Ads Help: About Performance Max campaigns). For retailers, it uses the Merchant Center feed alongside text, image, and video assets, and Google's automated bidding decides where and when to serve toward the conversion goal you set. Google's current documentation also describes controls that older guides said did not exist, including campaign-level negative keywords and brand exclusions, plus reporting on channels, search themes, and assets (same source). If you are reading an older onboarding page or agency guide that says Performance Max cannot use negatives, treat that statement as outdated and check the current help article.
The practical difference: Standard Shopping gives you a narrower tool with more direct levers. Performance Max gives you a broader tool whose levers are mostly inputs (goals, values, feed, assets, signals, exclusions) rather than direct placement choices.
The comparison, criterion by criterion
The table below compares both campaign types on the same criteria. It describes structural differences, not which will perform better for a given store.
Criterion Standard Shopping Performance Max --- --- --- Inventory scope Shopping placements on Google Shopping plus Search, YouTube, Display, Discover, Gmail, and Maps from one campaign Merchant Center feed Required; feed quality drives eligibility and matching Required for retail product ads; feed quality still drives product eligibility and matching Creative inputs Product data (titles, images, prices) only Feed plus text, image, and ideally video assets grouped into asset groups Goals and bidding Manual CPC or automated strategies such as target ROAS Automated bidding toward conversions or conversion value, optionally with a target Controls and exclusions Product groups, bids, negative keywords, search-term review Listing groups, campaign-level negatives, brand exclusions, URL and asset controls; no manual per-placement bids Reporting Search terms and product-level results in one channel Channel, search-theme, asset, and product reporting across mixed inventory Product and margin mix Easy to split products into campaigns or groups by margin tier Possible via listing groups or separate campaigns, but budget flows by the goal you give it Valid testing Clean baseline; can be the control in a campaign experiment Can be tested against Standard Shopping through Google's campaign-type experiment workflow
Two notes on the table. First, "more controls" is not the same as "better results"; controls only help if someone uses them against a clear economic target. Second, the reporting row matters for evidence: in Performance Max, one headline ROAS number can blend branded search, remarketing-like traffic, and new-customer prospecting across several channels. That blend is not wrong, but it is harder to interpret.
Why comparing the two campaigns in the interface misleads you
If both campaign types are live and eligible to serve the same product, they are not running a fair race. Google explains that Shopping ads can be served through either Standard Shopping or Performance Max, and when both campaigns include the same product, which campaign serves can be decided by Ad Rank (Google Ads Help: Performance Max and Standard Shopping campaigns). The same source points advertisers who want to compare the campaign types toward an experiment rather than reading side-by-side results.
In plain terms: when both campaigns target the same products, one campaign may win the auctions it was always likely to win (for example, high-intent branded or product searches) and look efficient because of that, not because the campaign type is superior. The other campaign gets what remains. Their ROAS figures reflect different slices of traffic. Any budget move made on that comparison is built on selection, not causation.
This overlap can lead teams to draw the wrong conclusion. Before using any "PMax vs. Shopping" result, ask: did both arms have access to the same products and the same demand, with traffic split by design rather than by auction?
Gate 1: conversion-value integrity and feed health
Both campaign types depend on the data you give them. Performance Max depends on it more, because it has fewer direct levers and relies on automated bidding toward the values you report.
Conversion values. Check that the purchase value Google receives matches what the business actually earns closely enough to steer bidding. Common gaps: values including tax or shipping, values sent before discounts are applied, duplicate purchase events, or test orders counted as revenue. Reconcile a sample period against your store platform's sales report. Shopify's sales report, for example, separates gross sales, discounts, returns, and net sales (Shopify Help Center: Sales reports). If the value you send to Google looks like gross sales but your margin depends on net sales after returns, bidding will overvalue high-return products.
Feed health. Product ads need an eligible, approved Merchant Center feed with accurate required attributes. Google's Shopping ads documentation and Merchant Center product data specification describe the required and recommended attributes, such as titles, prices, availability, images, and identifiers (Google Ads Help: About Shopping ads; Merchant Center Help: Product data specification). Disapproved products, mismatched prices, missing identifiers, and weak titles limit either campaign type.
Gate question: can you reconcile reported conversion value to store revenue for the same period, and is the feed free of material disapprovals and mismatches?
If no: fix the data first. Neither campaign type can reliably tell you which is better while the signal is broken. Standard Shopping is often the more interpretable holding pattern while repairs are made, because problems are easier to see at product and query level.
Gate 2: inventory and creative scope
Performance Max's broader inventory is only an advantage if you can supply what that inventory needs and if the business wants those placements.
Creative. Beyond the feed, Performance Max asset groups use headlines, descriptions, images, logos, and video. If you cannot provide relevant assets, the campaign can still run, but the non-Shopping placements will rely on thinner inputs, and you will have less to learn from asset reporting.
Inventory fit. Ask whether YouTube, Display, Discover, and Gmail placements match how your customers buy. A considered, visual product with a story may benefit from video and feed placements. A replenishment product bought on exact search intent may gain little from them.
Gate question: do you have, or can you produce, creative that matches your products and brand across video and image placements, and do you want spend on those placements?
If no: Standard Shopping keeps spend focused on Shopping placements where product data does most of the work. Performance Max remains an option once creative and inventory scope are ready.
Gate 3: product segmentation and margin economics
Platform ROAS treats a dollar of revenue as a dollar of revenue. Your business does not. A product with a high return rate, heavy shipping cost, or low gross margin can show strong ROAS while destroying contribution.
Segment products by economics. Group catalogue items by contribution margin tier (after cost of goods, fulfilment, payment fees, and expected returns). This is the same finance-owned margin contract described in our e-commerce contribution margin guide.
Decide how each campaign type will respect those tiers. In Standard Shopping, you can split products into product groups or separate campaigns and bid differently. In Performance Max, you can segment with listing groups or separate campaigns, or adjust the conversion values you send so they reflect margin rather than revenue. Either way, the campaign needs to know that not all revenue is equal.
Gate question: do you know contribution per order by product tier, and can the campaign structure reflect it?
If no: do not scale either campaign type on revenue ROAS alone. Build the margin view first; without it, any winner you declare may simply be the campaign that sold more low-margin, high-return products.
Gate 4: experimental feasibility
A more defensible way to answer "Performance Max or Standard Shopping for this catalogue" is a controlled comparison. Google offers a campaign-type experiment that compares a Standard Shopping campaign against Performance Max, with traffic split between the arms and guidance to use similar conversion goals and comparable bidding targets such as target ROAS (Google Ads Help: Performance Max experiments for Standard Shopping). Google's broader experiments documentation explains how experiments are set up and read, and notes limitations that apply (Google Ads Help: About Performance Max experiments; Google Ads Help: Performance Max experiments best practices). Check the current version of those pages before launching, because experiment options and requirements change.
Gate question: do you have enough conversion volume, a stable enough period, and the patience to run the experiment without changing both arms mid-test?
If no: do not pretend a before/after switch is an experiment. Seasonality, promotions, and feed changes will contaminate it. Keep the more interpretable setup, fix the preceding gates, and schedule the experiment for a stable window.
Disqualifying conditions for each option
These are conditions under which an option should not be the primary Shopping architecture right now. They are not permanent verdicts.
Standard Shopping is likely the wrong primary choice when:
The business deliberately wants reach into video and feed placements and has the creative to support it, and a valid experiment has shown Performance Max adds contribution.
The team has no capacity to manage product groups, bids, and search terms, so the extra control would go unused.
The catalogue and feed are strong, values are reconciled, and Standard Shopping has been the control in a test it lost on contribution.
Performance Max is likely the wrong primary choice when:
Conversion values cannot be reconciled to store revenue, or include tax, shipping, or pre-discount values the business does not keep.
The feed has material disapprovals or price and availability mismatches.
There is no margin view, so the campaign will optimise toward revenue that may not be profitable.
The business requires query-level control or needs to keep spend off non-Shopping placements.
There is no creative beyond product images and no plan to produce it.
If both lists describe your account, the problem is upstream of campaign type. Fix data and economics first.
An actionable experiment design
The following design keeps the comparison fair. Adapt it to Google's current experiment workflow.
Write the decision rule first. Example: "If Performance Max produces higher contribution after returns and variable costs at no worse new-customer share, shift the tested product set to it. Otherwise keep Standard Shopping." Deciding the rule after seeing results invites motivated reasoning.
Choose the product set. Use one product scope for both arms. Exclude products you cannot afford to test, such as limited stock or items about to change price.
Use Google's campaign-type experiment. Let the platform split traffic between the Standard Shopping control and the Performance Max arm rather than running two independent campaigns on the same products.
Match goals and bidding. Use the same conversion goals and comparable targets in both arms, as Google's workflow recommends. Different goals turn the test into a comparison of goals, not campaign types.
Freeze confounders. Avoid major feed rewrites, price changes, site redesigns, or promotions that affect only one arm. If a sitewide promotion happens, note it and consider whether it affects comparability.
Measure outside the platform. Pull orders by arm where possible, then join them to returns, discounts, and variable costs from your store and finance data. Google's reported values are one input, not the verdict.
Read the result against the pre-set rule. Report the contribution difference, the uncertainty around it, and what the test does not show (for example, effects on branded search or other channels).
A hypothetical contribution calculation
The numbers below are invented to show the method. They are not benchmarks, not expected results, and not from any client or account.
Assume one test period with the same product set in each arm.
Line item Arm A (Standard Shopping) Arm B (Performance Max) --- --- --- Ad spend $10,000 $10,000 Platform-reported conversion value $40,000 $46,000 Platform-reported ROAS 4.0 4.6 Revenue after discounts (store data) $38,000 $43,000 Returns (assumed rate) 10% → $3,800 18% → $7,740 Net revenue $34,200 $35,260 Cost of goods (assumed 45% of net revenue) $15,390 $15,867 Fulfilment, payment, and return handling (assumed) $4,000 $5,100 Contribution after ad spend $4,810 $4,293
In this invented scenario, Arm B shows higher platform ROAS but lower contribution because it sold more of a product mix with higher returns and handling costs. The assumptions carry the result: change the return rates or handling costs and the answer can flip. That is the point. Platform-reported ROAS is not proof of incremental contribution. It does not include returns, cost of goods, or fulfilment, and it does not show whether those sales would have happened anyway.
If your experiment only reports platform ROAS, you have learned which arm Google credits with more value, not which arm makes the business more money.
How this decision fits the rest of your Google spend
Shopping architecture sits beside other choices. Branded search, for example, can inflate both campaign types' results if brand queries are included; our guide to brand bidding incrementality covers that test. If the question is whether Google or Meta should get the next budget increment at all, see Google Ads vs. Meta Ads for e-commerce. And for the hierarchy between blended efficiency and platform metrics, see MER vs. ROAS.
The decision in one paragraph
Pass the four gates in order. If values or feed are broken, fix them and keep the more interpretable setup. If you lack creative or do not want non-Shopping placements, Standard Shopping fits the current scope. If you cannot see margin by product, build that view before scaling either option. If you can run a clean experiment, let Google split traffic between a Standard Shopping control and a Performance Max arm with matched goals, and decide on contribution after returns and variable costs, using the rule you wrote before the test began.
Want a second pair of eyes on your Shopping setup?
If you are choosing or reconsidering your Shopping campaign architecture, Sharply Labs can review your Merchant Center feed, conversion values, campaign overlap, and product economics, then return a written decision and test plan. We do not promise a specific improvement; the aim is a decision you can defend with your own data. Explore our e-commerce growth work or see how we run growth programs.