When Google Shopping Outperforms and When It Doesn't

· PPC

When Google Shopping Outperforms and When It Doesn't

Nonbrand Google Shopping beat nonbrand Search in 2 of the 4 ecommerce accounts we reviewed, tied in 1, and lost clearly in 1. In each account, the result came down to the query, the product, and the way the feed was built.

We pulled the last 30 days of Google Ads data from Data Druid, our internal account intelligence platform, for the 4 accounts running nonbrand Shopping. Two more accounts contributed figures for the branded comparison.

One caution applies to every number below. Blended Shopping ROAS overstates prospecting performance, because branded product queries sit inside the same channel. Those shoppers were already looking for you, so they convert at rates prospecting traffic won't match. Every comparison below separates brand from nonbrand for that reason.

What our accounts show

Nonbrand Shopping ROAS ranged from 1.41x to 5.56x across the 4 accounts. The account with the highest Shopping ROAS ran Search at almost exactly the same return.

Account (anonymized)

Approx. Shopping AOV

Nonbrand Shopping ROAS

Nonbrand Search ROAS

Nonbrand PMax ROAS

Shopping share of nonbrand spend

UK supplements brand

£22

5.56x

5.66x

Not running

61%

US consumer electronics brand

$110

3.17x

2.43x

2.23x

2%

US household essentials brand

$50

1.46x

1.27x

2.84x

22%

Personalized gifts brand

$90

1.41x

2.30x

1.84x

63%

Shopping is the supplements brand's largest nonbrand channel, holding a 5x+ return at that scale. The gifts brand also puts most of its nonbrand budget into Shopping, yet there it trails both Search and PMax by a wide margin.

Read the electronics and household numbers with budget share in mind. In the electronics account, the 3.17x comes from a small fallback campaign carrying 2% of nonbrand spend. Campaigns that size haven't been pushed into the pricier auctions that come with scale. In the household account, Shopping is restricted to new customers, so part of its lower return is by design.

When Shopping outperforms

Shopping wins when the listing itself can close the sale. The image, price, and product title give a shopper everything they need to choose.

The product sells on image, price, and a familiar category name

The supplements brand is the clearest case. Its single-product Shopping campaigns converted at 18% to 27% of clicks, with every one returning 4.79x or higher. Low-AOV, replenishable products with a familiar category name suit a format built for side-by-side comparison. A shopper searching for a common supplement already knows what they want, so the listing only has to win on price and trust.

The feed is split into campaigns by product line

The same account runs 8 nonbrand Shopping campaigns. Six cover core product lines, 1 covers bundles, and a catch-all holds lower-priority SKUs. Each campaign gets its own budget and bid target, so spend tracks demand across the catalog instead of pooling on the best seller.

Returns ranged from 4.79x to 6.97x across those 8 campaigns. The catch-all for lower-priority products posted the highest ROAS of the group while carrying 16% of nonbrand Shopping spend. That share is large enough that a small-budget effect doesn't explain it. One account sets no rule, though a separate long-tail campaign with its own bid target is a test worth running.

The query is your brand or your product name

Branded Shopping returned 3.75x to 70.3x across the 5 accounts running it. In 3 of the 4 accounts we could check side by side, branded Search returned more. Branded Shopping keeps your listings at the top of the results page when competitors or resellers bid on your name. It also captures demand that already exists, since these shoppers were looking for you. Keep it in its own campaign so its returns never prop up your prospecting numbers.

When Shopping underperforms

Shopping falls behind when the listing cannot carry the sale on its own. That happens when every listing looks the same, or when the product needs words to explain it.

The query is the literal name of the product type

The gifts brand splits its Shopping campaigns by query intent. The campaign built around the literal product-type query, the one most marketers would call highest intent, returned 0.71x. A campaign built around broader personalized-gift queries returned 1.83x on similar spend. An anniversary-gift campaign returned 2.03x on about a third of that budget.

When a shopper searches the exact product name, the results grid fills with near-identical listings at similar prices. In that grid, price and seller recognition decide most clicks. Broader gift searches give a distinctive product more room to stand out.

The product needs explaining

Personalized products are a hard fit for a format built on 1 image and a price. The gifts brand's nonbrand Search returned 2.30x against 1.41x for nonbrand Shopping. Ad copy can describe what a product image cannot show.

In the household essentials account, nonbrand PMax returned 2.84x, roughly double its Shopping and Search returns. Its Shopping campaign is restricted to new customers, so the gap overstates how far PMax is ahead. Before you credit PMax with a win like this, confirm both campaigns target the same customers. Check brand exclusions too, since PMax will pick up branded queries unless told not to.

High-ticket products trade conversion rate for order value

Considered purchases convert far less often in Shopping, while AOV can still carry the return. The electronics brand's standard Shopping campaign converted at 0.7% of clicks, with its Shopping-only PMax campaigns at 1.7% to 2.0%. Compare that with 18% to 27% for the supplements brand.

Order value closes much of that gap. The electronics fallback campaign returned 3.17x on a $110 AOV, ahead of the account's nonbrand Search. Expect fewer conversions per click on a considered purchase, then judge Shopping on revenue per click instead of conversion rate.

How to check your own account

Run these 5 checks before moving budget into or out of Shopping.

  1. Split brand from nonbrand. Report branded and nonbrand Shopping separately, then add brand exclusions to any PMax campaign you treat as prospecting.

  2. Compare nonbrand Shopping against nonbrand Search and PMax at similar spend levels. A small campaign with a high ROAS hasn't yet been tested in the pricier auctions that come with scale.

  3. Pull the search terms behind your weakest Shopping campaign. If the literal product-type query dominates, test separating it from broader gift, occasion, problem, and attribute queries.

  4. Ask whether your listing explains itself. For a product that is personalized, new to its category, or needs a demo, give Search and PMax creative a larger share of the test budget.

  5. Segment the feed by product line or margin tier. In a single all-products campaign, the best seller absorbs spend that long-tail SKUs could put to work.

Across these accounts, Shopping performed best where the listing could close the sale on its own. The supplements brand held a 5x+ return on 61% of its nonbrand budget. The gifts brand showed the opposite pattern, with Shopping taking 63% of nonbrand spend while trailing Search by nearly 1 point of ROAS. In both cases, the product and the query explained the result better than the channel did.

Start with the brand split this week. Then pull the search terms behind your weakest nonbrand Shopping campaign. If the literal product-type query dominates, separate it out and give Search and PMax creative a fair test at comparable spend. Move budget toward whichever channel holds its return as spend rises.

Methodology

All data is anonymized. Figures come from Google Ads campaign data pulled through Data Druid on October 6, 2026, covering the previous 30 days. Conversions and revenue are reported by conversion time using Google's own attribution. Channel ROAS is spend-weighted across all campaigns of each type, with Shopping-only PMax campaigns counted under PMax. Brand versus nonbrand follows each account's campaign classification.

Four accounts ran nonbrand Shopping in the window. Two more contributed branded Shopping figures only. Currencies differ by account, so AOV is approximate and ROAS is the comparable measure. This is a small, single-window sample, best read as directional evidence rather than a benchmark.


Back to all articles