Standard Shopping is still the campaign type that gives you the most direct say over which products get budget, what you bid on them, and which queries you pay for. For a long time, that control was the main reason to choose it over Performance Max.

The gap is narrower now. PMax supports up to 10,000 campaign-level negative keywords, shared negative keyword lists, and search terms reporting. And in 2026, Google brought AI Max to Standard Shopping, adding AI-written ad text and broader query matching to the campaign type that used to be the manual option.

So the question is less about which campaign type to pick and more about how to structure Shopping campaigns so they work alongside PMax and AI Max instead of competing with them.

Below are six structures, when each is worth its upkeep, and what AI Max changes about query sculpting in each.

The three decisions that shape a Shopping structure

Shopping campaigns have four levels: campaign, ad group, product group, and product. Most structural decisions, though, come down to three questions: where you set priority, where your negatives live, and which products share a product group’s bid and budget. Each level controls a different part of that.

Campaign level

Campaign settings apply to everything underneath them: the ROAS or CPA target, budget, bid strategy, campaign priority, location and language targeting, audience segments, and AI Max for Shopping. Of these, priority and AI Max have the most influence on structure, and both get their own sections below.

If you use Target ROAS or Target CPA, set a target your conversion data can support. Smart Bidding needs clean conversion data and enough volume to bid well. A campaign with a handful of conversions a month and an aggressive tROAS target will often stop spending, because the system can’t find enough auctions it expects to meet the goal.

Ad group level

A Shopping campaign can hold multiple ad groups, and negative keywords are what make that useful. You can add negatives at the ad group level, at the campaign level, or through shared lists.

This is still one of Standard Shopping’s biggest practical advantages. PMax supports campaign-level negatives and shared lists, but they apply only to its Search and Shopping inventory, and you can’t split one PMax campaign’s traffic between product groups the way separate ad groups let you in Standard Shopping.

That control only pays off if someone reviews search terms regularly for irrelevant or expensive queries that don’t convert. Optmyzr’s Rule Engine can take on part of that work by flagging underperforming queries based on aggregated performance within each ad group. You can also describe the rule in plain language (in beta) and have Optmyzr draft a strategy for you to review before it runs.

Product group level

Product groups sit inside ad groups, and you can build them from two kinds of data. Feed attributes such as brand, product type, category, and color work when those attributes line up with how you want to bid. Business data such as profit margin, bestseller status, and seasonality usually isn’t in the feed by default, so it’s passed in through custom labels.

Product groups are where you set bids in manual CPC campaigns and where you decide which products share a budget. That makes them the level where business goals turn into bidding decisions. Building product groups from custom labels lets you push high-margin items, give seasonal collections their own budget, or move clearance stock without it competing with full-price products.

Some advertisers go a step further and move products between groups automatically as performance or inventory changes. That keeps budget on what’s selling, but it only works if something maintains the labels. Done by hand, it falls behind within weeks.

Campaign priority

Each Standard Shopping campaign can have a priority of high, medium, or low. When the same product sits in several Standard Shopping campaigns, priority decides which campaign’s bid enters the auction.

Priority is what makes query sculpting possible. The classic setup runs three campaigns containing the same products, each with a different job:

• High priority, low bids: catches broad, generic searches cheaply
• Medium priority, medium bids: catches category and brand searches
• Low priority, high bids: catches specific product searches, where intent and conversion rates are highest

The setup depends entirely on negatives. Without them, the high-priority campaign catches every query, because its bid is the one Google uses. Adding brand and product terms as negatives in the high-priority campaign pushes those searches down to the medium and low campaigns. The medium campaign then gets product-specific negatives, so the most specific searches fall through to the low-priority, high-bid campaign.

6 ways to structure Google Shopping campaigns

There’s no single correct structure. The right one depends on your catalog size, budget, margins, and how much time your team has to maintain it.

Each option below adds control, and each adds upkeep in roughly the same proportion.

1. Single campaign, single ad group and product group

Suits brands selling one product line at similar prices, such as flavored sparkling water.

• Works for accounts with small budgets and small teams
• Gives Smart Bidding the most data in one place
• Hard to see product-level performance once the catalog grows
• Hard to set different bids or negatives for individual SKUs

2. Single campaign, multiple ad groups and product groups

Suits brands with one product line and a wide price range, such as footwear.

• Lets you split products by category, brand, margin, or price tier (under $50, $50 to $150, over $150), so you can react to pricing changes without restructuring
• Needs more negative keyword management to keep ad groups from overlapping
• Budget is still shared, so it’s hard to push more spend to your best products

3. Two or three campaigns using campaign priorities

Suits brands with a limited but varied product line that have outgrown a single campaign.

• Gives you more control over where budget goes
• Priority tells Google which campaign’s bid to use for a given product
• Use high priority for broad, low-intent searches; medium for category or brand searches; low for specific, high-intent product searches where ROAS is the goal
• Needs negatives on the high and medium campaigns, or the high-priority campaign catches everything
• Needs enough budget that each campaign can run without constantly hitting its cap

4. Multiple campaigns, multiple ad groups, and product groups with query sculpting

Suits large retailers, resellers and accounts with many product types and brands.

• Gives the tightest budget control
• Uses negatives and priorities together to send each query to the right campaign
• Hard to maintain by hand as feed values change, stock runs out, and product lines are added or dropped
• Overspends or misroutes traffic quickly when nobody’s watching

At this scale, manual upkeep usually falls behind the feed.

You need negative keyword lists that update with search term data and campaigns that resync when the feed changes.

5. Performance-based structure

Suits advertisers who need tight control over spend by product.

• Groups products by performance (ROAS, CPA) or business metrics (production cost, contribution margin)
• Moves budget to the products that return the most
• Takes testing to find the right target for each group
• Hard to build and maintain without automation, because products move between groups as their numbers change

Margin-informed bidding makes it stronger. Passing margin or cost data into custom labels, or reporting conversion value as profit instead of revenue, points Smart Bidding at the sales that make money rather than the ones that only look good on revenue.

Optmyzr’s Smart Product Labeler handles the grouping by sorting products into five buckets based on ROAS and cost: heroes (high return), sidekicks (close to hero level), villains (high cost, low return), zombies (too little traffic to judge), and flukes (unusually high returns on little exposure).

Each bucket can then get its own campaign or target. The tool can also factor in feed attributes, and it generates a supplemental data source you add in Merchant Center, so the labels update on a schedule without scripts.

6. Groups of individual products (GRIP)

Suits catalogs with a handful of high-value products, or new launches you want to isolate.

• One product per product group, for the most control over bids
• Best for small catalogs (single-digit or low double-digit SKUs) or a handful of high-value products
• Useful for testing new products and collecting clean performance data
• Splits conversion data thin, so it works best with manual CPC bidding, not tROAS

How to choose between them

Start with the simplest setup your catalog and budget allow, and add structure only when a specific problem shows up that the current setup can’t handle.

A single campaign is the right default while conversion volume is low, because it keeps data together for Smart Bidding. Move to multiple ad groups when products need different negatives or bids. Move to a two- or three-campaign priority split when your search term labels show clearly different intent tiers and you have the budget to keep each campaign from capping out.

Full query sculpting (setup 4) earns its maintenance cost when the catalog spans many product types and brands. A performance-based split (setup 5) works once each bucket has enough conversions to hold its own target. GRIP sits outside this progression: use it for the few products that justify individual attention.

Splitting products between Standard Shopping and PMax

Many accounts now run Standard Shopping and PMax side by side. Because Ad Rank decides between them on overlapping products, the cleanest hybrids split the catalog so each campaign type has a clear job.

For example, you might run PMax on heroes and sidekicks, where it has enough data to scale, and keep villains, zombies, and new products in Standard Shopping, where you want manual control and full search term visibility.

Custom labels make that split maintainable, but only as long as something keeps the labels current. If a hero drops to villain and its label doesn’t change, it stays in PMax with a budget it no longer earns.

Read more: 3 Google Shopping Campaign Examples You Can Copy Now

What AI Max for Shopping changes

Google announced AI Max for Shopping on April 30, 2026 as a one-click upgrade for Standard Shopping campaigns, and it started appearing in advertiser accounts in late July. Google’s help documentation still labels it a beta. It’s aimed at the longer, conversational queries that product-title matching tends to miss, including searches in AI Overviews and AI Mode.

It adds three features to a Standard Shopping campaign:

• Text customization: Google rewrites product titles and ad text to match the shopper’s query, using attributes from your Merchant Center feed. It currently works only with English-language feeds.
• Final URL expansion: Google can send clicks to whichever page on your site it predicts is most relevant, not only the product page. It requires text customization to be on.
• Optimal format selection: Google chooses whether to show a Shopping ad or a text ad for a given query.

Your product groups, priorities, and bid strategy stay as they are. You can also turn off final URL expansion to keep delivery limited to Shopping ads.

What it means for your structure

• Test one campaign first. Standard Shopping keeps its structure under AI Max, so you can switch it on for a single campaign and compare. A reasonable starting configuration is text customization on and final URL expansion off. Final URL expansion only works with a Target ROAS bid strategy.
  
• Wait before judging. Give it at least two weeks, and look at account-level results, not only the AI Max campaign. New queries it picks up can shift volume between campaigns.
  
• Watch your sculpting. AI Max widens the queries a campaign can match. In a priority-based setup, that means more traffic can land in your high-priority, low-bid campaign. Check search terms weekly and add negatives where traffic lands in the wrong tier. Campaign priority applies to Shopping ads and text customization, but not to final URL expansion.
  
• Your feed matters more. Text customization writes from your feed attributes. Thin descriptions and missing attributes (material, size, color, fit) give Google less to work with. Fix the feed before you turn it on.
  
• Check the product titles report. It shows what Google rewrote, so you can catch claims or wording you don’t want on your ads.

Google’s own guidance goes further: it recommends turning on both text customization and final URL expansion, judging results at the account level, and not chasing every new match with negatives.

That’s reasonable for a single-campaign setup. In a priority-tiered structure, though, loose queries landing in the high-priority campaign can undercut the whole sculpt, so start narrower and widen once you’ve seen where the new traffic lands.

If you already run Broad Match Search and full PMax, AI Max for Shopping may add little you aren’t already reaching. It’s most useful for accounts that rely on Standard Shopping and exact-match Search. Treat it as a test with a clear success metric, not a default setting.

Keeping your Shopping structure working

A structure holds up only as long as the data and settings around it stay accurate. How much each area below matters depends on your catalog, budget, and team, so treat these as questions to ask of your own account rather than a checklist to complete.

Product grouping

Group products the way you actually make decisions. Feed attributes like brand and color work when shoppers search with those terms. Category and subcategory groups make reporting easier. Grouping bestsellers together helps move inventory, and grouping by margin protects profit.

For finer control, build product groups from custom labels tied to data that changes: margin, seasonality, stock level, or performance bucket. Google allows five custom labels per product (0 to 4) and up to 1,000 unique values per label. If you manage more than one feed in a Merchant Center account, keep labels consistent across feed labels so the same rule means the same thing everywhere.

The practical difficulty is keeping campaigns in step with the feed. Optmyzr’s Shopping Campaign Management tool builds Standard Shopping and PMax retail campaigns from your Merchant Center feed, based on feed attributes or product performance, and resyncs them on a schedule. If your product groups run on performance labels, Smart Product Labeler (see setup 5) keeps those labels current as the numbers change.

Merchant feed quality

Your feed decides what Google can show and how well it matches queries. With AI Max writing ad text from feed attributes, it now shapes your ad copy as well.

The common problems haven’t changed much:

• Titles that are too long, too vague, or missing the terms people search for
• Missing attributes, especially GTIN, brand, color, size, and material
• Low-resolution or unappealing images
• Disapproved or inactive products

Three Merchant Center changes in 2026 add to that list:

• The image minimum is rising to 500 x 500 pixels. Merchant Center has warned about undersized images since April 14, 2026, and enforcement starts January 31, 2027 for both image_link and additional_image_link. Undersized images show up in the Needs attention tab.
  
• Product videos are eligible to serve. The optional video_link attribute has been eligible to serve since June 30, 2026. A video with policy or quality errors won’t serve, but it won’t block the product either.
  
• The Content API for Shopping has been sunset. Google sunset it on August 18, 2026, and requests without an extension have been failing intermittently since September 1, 2026. Any feed tool or integration still using it needs to move to the Merchant API. File uploads, scheduled fetches, and Google Sheets feeds aren’t affected.

Google also added product-level shipping attributes in 2026, including handling cutoff time and minimum order value. They help when some brands or suppliers ship on different terms from the rest of your catalog.

Finding these problems across a large feed is slower than fixing them.

Optmyzr’s Shopping Feed Audits tool grades your feed along with your Shopping and PMax retail campaigns. It flags missing data, disapproved products, oversized product or listing groups, campaigns without negatives, and pricing gaps against benchmarks. You can fix issues from the results page by hand, in bulk, or with AI suggestions, and schedule audits to run and email you the results.

For larger fixes, the new Feed Optimization (beta, Google Ads only) lets you edit titles, descriptions, brands, prices, links, and images in bulk, either with AI suggestions you review or with rules that apply fixes automatically.

Changes go to a supplemental data source in Merchant Center, so your primary feed stays untouched, and automated refreshes keep the fixes in place as the feed updates.

To monitor feed and product health at scale, the Rule Engine can alert you to, or act on, disapprovals, underperforming product groups, and products that are trending up or losing conversions.

Budgets and bid strategy

Budget decisions carry more weight in Shopping than in most campaign types, because shoppers compare you with competitors on the same results page.

• Match structure to budget. Brands succeed with Shopping whether they spend $5,000 a month or $50,000 a day. But a small budget split across many campaigns and product groups creates internal competition and keeps each one from entering enough auctions to learn.
• Pick a bid strategy for your goal. Standard Shopping gives you control over structure, and Smart Bidding often beats manual bidding once there’s enough data. If you can only pay a certain amount per sale, a CPA target gives Google a clear limit. If revenue matters more, tROAS fits.
• Bid on margin, not only revenue. Custom labels can separate products by margin, or you can report profit as conversion value, so Smart Bidding favors the sales that make money.
• Total budgets for fixed flights. Since January 15, 2026, campaign total budgets have been in open beta for Search, PMax and Shopping. They work for campaigns with set start and end dates, so a sale or launch doesn’t need daily budget changes by hand.
• The world outside the account. Seasonality, promotions, competitor pricing and stock levels all move performance. Targets and budgets set ahead of known peaks such as Black Friday hold up better than ones changed mid-peak.

Optmyzr’s Budget Monitors pace spend against monthly or custom budget cycles, can auto-adjust or pause, and can roll unspent budget into the next cycle.

Negative keywords

Shopping campaigns don’t use keywords for targeting, but they do accept negatives, and in a sculpted structure the negatives do most of the work. A negative added to an ad group stops every product in that ad group from showing for that search.

Each level has a different job. Ad group negatives do the sculpting, campaign negatives handle exclusions for one campaign, and shared lists or account-level negatives cover terms you never want to pay for, such as “jobs,” “free,” or “repair.” Negatives use the same match types as regular keywords (broad, phrase, and exact) but don’t match close variants, so you need to add plurals and synonyms yourself.

Regular search term reviews, weekly for larger accounts, catch expensive or irrelevant terms before they use up budget and surface new queries worth routing to a different campaign. With AI Max for Shopping on, more frequent checks help at first, since it matches a wider range of queries than the campaign did before.

For PMax campaigns running alongside Standard Shopping, search terms reporting, campaign-level negatives, and shared negative keyword lists are all available. That makes it easier to see whether the two campaign types are catching the queries you intended.

Conversion tracking

The feed decides what Google can show. Conversion tracking decides what Smart Bidding learns from, because the actions you mark as primary are the ones it optimizes toward.

When the data is incomplete or wrong, the account learns the wrong lessons, and Smart Bidding and AI Max act on those lessons across every auction they enter. A tracking error that would be a minor reporting issue in a manual account becomes a bidding problem in an automated one.

End-to-end validation means checking that tags fire correctly, that consent mode is configured consistently, and that conversion values match what your store records. Enhanced conversions help recover sales that browser restrictions would otherwise hide, and in 2026 Google merged the web and leads versions into a single on/off setting.

Audience segments

First-party segments act as bidding signals for Smart Bidding without restricting reach. The more Shopping-specific lever is the new customer acquisition goal, which works in two modes: bid higher for new customers, or bid only for new customers. Segmenting by customer stage (new visitors, cart abandoners, and repeat buyers) shows which mode fits.

Read more: How to optimize Google Shopping campaigns

Keeping it all in sync with Optmyzr

Picking a structure is a one-time decision. Keeping it working is ongoing: feeds change, stock runs out, queries drift between priority tiers, and every campaign type you add is another place for that drift to happen. The feed and campaign-building tools are covered above. For routing queries and setting bids, these help:

• Routing queries to the right ad group. Negative Keywords Finder for Shopping compares how the same query performs across ad groups and adds it as an exact-match negative where it underperforms, so the traffic lands where it converts best. Search Term N-Grams shows which words cost money without converting. For PMax, the PMax Search Terms and Non-Converting Queries - PMax strategies in the Rule Engine surface the queries worth excluding.
  
• Bidding by attribute, not only by structure. Shopping Attribute Bidder aggregates performance by feed attributes such as price, color, or category and adjusts bids regardless of how your product groups are built. Shopping Analysis shows where performance clusters, such as the price points with the best ROAS.
  
• Working through it in plain language. Sidekick analyzes search terms, summarizes performance, and helps build Rule Engine strategies, and it shows every change for review before it goes live. Through Optmyzr’s MCP integration, you can do the same from AI assistants such as Claude and ChatGPT.

Already on Optmyzr? Open Shopping Campaign Management to build this structure from your feed. New to Optmyzr? Start a 14-day free trial today.

FAQs

1. What’s the best structure for a small product catalog?

For small catalogs, one campaign with one or a few ad groups and product groups is usually enough. It keeps conversion data together so Smart Bidding can learn. Label search terms by intent to decide where negatives are needed, and move to a two- or three-campaign priority setup once you have the volume and budget to support it. For a handful of high-value products, GRIP gives you bid control per product.

2. Can I still use Standard Shopping alongside PMax?

Yes, and many advertisers do. Since October 2024, PMax no longer takes automatic priority over Standard Shopping in the same account. When both target the same product, the campaign with the higher Ad Rank serves. The cleanest setups split products between the two, often with custom labels, so each campaign has a clear job.

3. What is AI Max for Shopping, and should I turn it on?

AI Max for Shopping is a Google Ads beta that adds text customization, final URL expansion, and format selection to Standard Shopping campaigns while keeping your product targeting and bidding. It’s most likely to help if you rely on Standard Shopping and exact-match Search and want to reach longer, conversational queries. The way to find out is a single-campaign test with text customization on and final URL expansion off, judged at the account level after at least two weeks.

4. How do I keep my Shopping structure up to date as inventory changes?

Moving products between campaigns by hand as performance or stock changes is slow and error-prone, so it’s worth automating once the catalog grows past what you can review manually. Optmyzr’s Shopping Campaign Management resyncs campaigns with Merchant Center on a schedule, and Smart Product Labeler updates the custom labels your structure depends on.

5. What feed issues hurt Shopping campaign performance?

The most common ones are:

• Long, vague, or keyword-poor product titles
• Missing GTINs or other required attributes
• Low-resolution images (Google’s 500 x 500 pixel minimum is enforced from January 31, 2027)
• Disapproved or inactive products
• Integrations still using the Content API for Shopping, which Google sunset in August 2026

6. How do campaign priority settings affect performance?

When the same product sits in several Standard Shopping campaigns, priority (high, medium, or low) decides which campaign’s bid enters the auction. That lets you bid low on broad, top-of-funnel traffic in the high-priority campaign and higher on specific, bottom-of-funnel searches in the low-priority one. It only works with negatives that push specific queries down to the lower-priority campaigns. Priority doesn’t apply between Standard Shopping and PMax; Ad Rank decides there.