Products Not Showing in Google Shopping: Which to Bring Back

Most products with zero impressions have sound data. In our study of 1.4 million products across 130+ stores, 55.0% received no ad budget over 13 months. A product stays unseen for one of four reasons: no campaign includes it, Google can’t show it, a few bestsellers take the budget, or it never got a fair test. Bring back products no campaign includes, untested ones, and disapproved ones you sell after a data fix. Leave out what you can’t sell: out-of-stock products until they return, discontinued ones for good.

How many products get no budget or impressions?

In our study of 1.4 million products (1,404,808 products, 130+ stores, 13 months), more than half of all products got no ad budget at all. The study split each catalogue by what the ads did with it:

An ordinary report hides this part of the catalogue. A performance report shows what happened: spend, clicks, sales. A product with zero impressions has nothing in those columns, so it rarely surfaces as a problem. The number that shows it is assortment coverage: the share of the catalogue that takes part in advertising.

Our guide to measuring catalogue coverage shows how to calculate your own coverage and what to compare it with.

Left alone, most of the no-budget part stays unseen. Of 1,196,357 products that started the period with zero spend, 74.1% never received a single impression in 13 months. Only 2.59% reached at least one conversion. Those that did took a median of 4 months to their first sale.

In most stores, these products sit inside Performance Max. It runs in 141 of 146 stores in our data and takes a median 95.7% of their ad budget (GetProfit data, June 2025 – June 2026). Our Performance Max guide for online stores explains how the campaign type works as a whole.

One date matters if you read product-level impressions in Performance Max. From 15 June 2026, the product report in the Google Ads API (the Shopping Performance View) includes all Performance Max networks, such as YouTube. It has always included the Display Network (Retail campaign performance, Google Ads API documentation). Google’s developer blog adds that API requests for earlier dates don’t contain the expanded data (Product reporting changes for Google Ads starting June 15, 2026). So a product with zero impressions in an older API report may still have appeared on a network that report didn’t cover.

Why do products get no impressions? The four groups

A product gets an impression only after it passes three gates: Merchant Center approves it, a campaign includes it, and it wins auctions within your budget. Our guide to how Google Shopping ads work follows that path step by step. Every unseen product is stuck at one of the gates. The last gate fails in two different ways, so there are four groups.

Groups 1 and 2 leave a trace in Google’s product statuses. In groups 3 and 4 the statuses look clean: the product is eligible and included, and still gets nothing.

Feed errors explain few of the unseen products

We checked which issues Google attaches to products across our stores’ catalogues. These are observations, not an experiment.

Issues cluster in a few catalogues. Google attaches at least one issue to 55.9% of the 5,744,507 products in our stores’ catalogues. Yet among the 142 stores with at least 100 products, the median store has 0.0% of its catalogue flagged. In 109 of them, fewer than 1% of products carry any issue (GetProfit data, catalogue snapshots in a data bank collected in August 2026). A handful of very large catalogues hold most of the flags.

The leading issues are decisions, not data errors. Among 400,000 flagged products (the first our query read, not a random sample), 263,483 carried “not eligible in any campaign” and 209,328 “out of stock”. No shipping set for the country came third, with 38,032. Classic data errors, such as a missing required attribute or an image that is too small, counted in the hundreds in that sample. The first issue is a campaign setting, the second a stock decision.

In most stores, the unseen products look healthy. Where the issue list is almost empty, Google’s statuses show an unseen product as eligible and included. It is crowded out or was never tested (groups 3 and 4), so the fix lies in budget and testing rather than in the feed.

Group 1: products no campaign includes

A Performance Max campaign takes products from the Merchant Center account named in its settings and can limit them to one feed label. If you leave the feed label empty, the campaign uses all feeds in that account (Performance Max for online sales with a product feed). Inside the campaign, listing groups decide which products each asset group includes or excludes (Listing groups for retail). The same page states: “Ads won’t serve until you create a valid listing group tree.”

A product can end up in no campaign in several ways:

• a branch covering its brand, product type or custom label is set to Excluded;
• the catch-all “everything else” branch is excluded, and the product’s attribute value changed in the feed, so it now lands there;
• the campaign is limited to one feed label, and the product carries another.

The listing groups guide adds that in Performance Max they work best for groups of products, split by dimensions such as custom labels or brand, rather than for single items.

Our article on listing groups in Performance Max helps you choose which attribute to split by. Our guide to the “not eligible in any campaign” issue shows how to clear it.

Bring these products back first. Their product data is already in order: one setting returns them to the campaign.

Group 2: products Google can’t show

Google Ads gives every product one of three statuses: eligible, eligible with limits, or not eligible. A not-eligible product cannot show in ads. Each product also carries a list of issues that affect whether it can show. Both the status and the issues can take up to 24 hours to update after a change (ProductStatus and ShoppingProduct in the Google Ads API reference).

The first reason is a failed approval: Merchant Center has disapproved the product. The product’s issue list names what blocks it, and the fix sits in the product data rather than in the campaign.

Stock is the second reason. “Out of stock” is the second most common issue among the 400,000 flagged products in our data. Keep availability accurate, and for other cases use the attributes Merchant Center provides (product attributes in the Merchant API reference):

What to bring back here: disapproved products you sell, once you fix their data. What to leave out: out-of-stock products until they are back in stock, and discontinued products for good.

Group 3: products crowded out by a few bestsellers

These products are eligible and included, yet the budget goes elsewhere. In our study, the median store spent 27.7% of its budget on its top 1% of products and 68.1% on its top 10%. The other 90% of the catalogue shared about a third.

In Performance Max, listing groups only include or exclude products; unlike in Standard Shopping, they don’t set bids (Listing groups for retail). Each Performance Max campaign has its own budget, which it can’t share with other campaigns (Create a budget, Google Ads API), so every product in the campaign draws on it.

In our data, the products that already sell take most of that budget, and the rest of the catalogue shares the remainder. The budget is always spread unevenly; when budget concentration starts to hurt is a separate question.

Look inside this group for products that shoppers click when they do see them. A high click-through rate (CTR) on a handful of impressions is a hint, not proof: small numbers swing. Our guide to high-CTR products with few impressions shows how to find these products and test them without fooling yourself.

Bringing them back takes a budget decision, such as moving them to a separate campaign with its own budget, rather than a settings fix.

Group 4: products that never got a fair test

A product with a few clicks and no sale is still an open question. Our study puts the minimum honest test at 15–20 clicks, about $3–4 at the sample’s average cost per click of $0.20. Below that, conclusions come from noise.

Unseen products that do start selling take their time. In the study, products that started with zero spend and later sold took a median of 4 months to their first sale. More than half of them went through a “clicks but no sales” stage first.

Keep this group apart from zombie products: a zombie gets impressions and clicks and spends money without a sale, while a never-tested product has spent nothing. Only products with spend give you performance to judge. For why “zombie” means different things to different people and how often such products come back, see zombie products explained.

A common answer for this group is a separate campaign for unseen products. Before you build one, read what a zero-impression campaign delivers. Either way, size the test in clicks, not in products: the number of products times 15–20 clicks, times your cost per click.

Approved but no impressions: what to check, in order

When a single product looks approved and still gets nothing, go through these checks in order. Each one rules out a cause before you change anything.

1. Make sure the zero is real. Check the date range first. For dates before 15 June 2026, remember that the product report in the Google Ads API didn’t cover every Performance Max network.
2. Read the product’s status in Google Ads. You will see eligible, eligible with limits or not eligible, plus a list of issues. After any fix, allow up to 24 hours before you judge.
3. If it is not eligible, open Merchant Center. Read the issue: disapproval, availability, shipping for the country. Fix the data or the stock first: a product Google can’t show needs that fix before any campaign setting can help.
4. If it is eligible, check that a campaign includes it. Compare the campaign’s Merchant Center account and feed label with the product’s. Then open the listing groups of each asset group in the Performance Max campaign and look for an excluded branch that covers the product’s brand, product type or custom label.
5. If the whole campaign barely gets impressions, the problem sits above the product. Work through why Performance Max is not spending before you look at single products.
6. If the campaign runs normally and only this product gets nothing, it is crowded out or was never tested. That calls for a budget decision rather than a settings fix.

Which unseen products to bring back, and which to leave out

Products that got their test, spent money and never sold are a different case. These are products that spend without selling, and the decision about them is whether to exclude them.

In our study, switching such a product off was risky at every spend level. The study tested the rule “no sale after a given spend, switch it off”. For products with $5 or more of spend, the rule was right 88.1% of the time; for $500 or more, only 66.2%. And 35.4% of products that had spent at least the cost of one conversion without a sale went on to convert, a median of 2 months later.

Our article on when to exclude a product shows how to set a limit tied to your margin. Our guide to excluding products from Performance Max compares listing groups, the excluded destination attribute and the pause attribute.

A worked example: sorting 1,800 unseen products

Example store, not client data.

A tableware shop has 3,000 products in its feed. It spends 40,000 a month, buys 8,000 clicks at an average cost per click of 5 and gets 180,000 in revenue. Only 1,200 products got an impression this month, so 1,800 products, or 60% of the catalogue, stayed unseen. Here they are, sorted into the groups:

Two of these fixes cost no ad money. Including the brand returns 450 products to the campaign, and fixing the price data returns 50 more. If all of them then get impressions, coverage grows from 1,200 to 1,700 products, or from 40% to 56.7% of the catalogue.

The 1,200 included but unseen products are the expensive part. A test at the lower end, 15 clicks each, needs 18,000 clicks. At a cost per click of 5, that is 90,000: 2.25× the monthly budget, and 2.25× the 8,000 clicks the shop buys in a month.

So the shop tests in waves. A first wave of 100 products from categories that already sell needs 1,500 clicks, which cost 7,500, or 18.75% of a month’s budget.

What to do with your own catalogue

1. Count the unseen products, starting from the full product list. A performance report lists only products that got impressions, so the unseen ones are easy to miss there.
2. Sort them into the four groups by status and issue: in no campaign, not eligible, included but crowded out, never tested. The groups need different fixes, so the counts tell you where to start.
3. Fix inclusion first. It costs no ad money and returns products to the campaign at once.
4. Fix disapprovals on products you sell, and keep availability accurate. Remove discontinued products instead of marking them out of stock.
5. Plan tests in clicks. Multiply the number of products you want to test by 15–20 clicks and by your cost per click before you start.
6. Check again after a day and after a month. Google needs up to 24 hours to update a product’s status, and a test needs its clicks before it tells you anything.

The portal’s Assortment section runs this count for the whole catalogue, with impressions, spend and sales over 90 days. It walks every product through the stages from the feed to sales: In the catalogue, Eligible for ads, Showing in ads, Spending budget and Bringing conversions.

Products whose only problem is that no campaign includes them count as ready for ads. The portal keeps two reasons for zero impressions apart: in no campaign at all, or in a campaign but never shown. It also names the categories where nine out of ten ads-ready products got not a single impression over 90 days.

See which part of your catalogue works. Sign in with Google in one click. The portal changes nothing without your consent.

Break down your catalogue →

Sources

• ProductAttributes — Merchant API reference: the availability, pause and excluded destinations attributes; a pause for ads covers all ads locations, including Shopping ads. Checked 2 October 2026.
• Performance Max for online sales with a product feed (retail) — Google Ads API: Merchant Center account and feed label in the campaign; without a feed label the campaign uses all feeds. Checked 2 October 2026.
• Listing Groups for Retail — Google Ads API: listing groups include or exclude products in each asset group and don’t set bids; they work best for groups of products, such as by custom label or brand; every subdivision has an “Other” node; ads won’t serve without a valid tree. Checked 2 October 2026.
• Create a budget — Google Ads API: a Performance Max campaign needs its own budget, which can’t be shared. Checked 2 October 2026.
• ShoppingProduct — Google Ads API reference: how a campaign includes a product; status and issues can take up to 24 hours to update. Checked 2 October 2026.
• ProductStatus — Google Ads API reference: eligible, eligible limited, not eligible. Checked 2 October 2026.
• Retail Campaign Performance — Google Ads API: from 15 June 2026 the Shopping Performance View includes all Performance Max networks. Checked 2 October 2026.
• Product reporting changes for Google Ads starting June 15, 2026 — Google Ads Developer Blog, 30 April 2026: requests for earlier dates don’t contain the expanded data. Checked 2 October 2026.
• GetProfit study of 1,404,808 products, 130+ stores, 13 months — catalogue split by spend and sales; products that started with zero spend; budget concentration; minimum test; accuracy of switching off; later conversions.
• GetProfit data: catalogue snapshots in a data bank collected in August 2026 — share of products with issues (5,744,507 products); share of the catalogue flagged in the 142 stores with at least 100 products; most common issues among 400,000 flagged products (not a random sample).
• GetProfit data: 146 stores, June 2025 – June 2026 — Performance Max share of stores and of ad budget.