Review sources

How to show Google Shopping seller ratings on your website

9 September 2026 · JustReview

Google Shopping seller ratings can be pulled onto your own website by a widget, but they are not the reviews most shop owners think they are connecting. A seller rating is a score of your shop: did the order arrive, was it what was promised, was the service acceptable. It says nothing about whether the item itself is any good. In our production database that separation is absolute, and knowing it changes where on the page these belong.

All figures below were measured in our production database on 9 September 2026.

Every single one is a seller rating, none is a product review

We hold 16,158 Google Shopping reviews across 12 accounts. Checking whether each one is attached to a product gives an unambiguous answer: zero are flagged as product reviews and zero carry a product title. Not a majority, not most. None.

That is not a gap in our import, it is what the source is. Google Customer Reviews collects a post-purchase survey about the order, and the star rating it produces feeds the seller rating that can appear next to your shop in Shopping results. Product reviews are a separate Merchant Center feed with a separate lifecycle, and they never arrive through this connection.

The practical consequence is placement. A seller rating is a shop-level trust signal, so it belongs where the buyer is deciding whether to trust the shop at all: in the header, near the checkout button, in the footer, on a delivery-and-returns page. Putting it on a product page under the heading “customer reviews” is a quiet misrepresentation, because a visitor reading four hundred five-star ratings under a specific item will assume those ratings are about that item.

The rating distribution has a hole in the middle

Here is how Google Shopping compares to ordinary Google Business reviews. We used reviews dated August 2026 for the Google column so that the comparison is against a recent, full month:

The headline share of five stars is similar. The shape underneath is not. Google Shopping has 1,209 four-star ratings against 97 three-star ratings, a twelve-to-one gap between two adjacent scores, and only 80 one-star ratings out of 16,158. A survey shown to a customer who just received their parcel gets answered by people whose parcel arrived. Everyone else closes it.

There is a second difference worth designing around. Where the review has text at all, it averages 63 characters against 191 on Google Business. Sixty-three characters is “Everything arrived on time, well packed.” These are useful as a volume signal and weak as testimonials, which is why a summary or counter layout suits this source better than a wall of quote cards. Our own recommended layout for this connection is the rating summary, not the review list.

Twelve accounts, and one of them is three quarters of the source

Fifteen accounts in our database have a Google Shopping connection configured. Twelve of them hold any reviews at all, so three, one in five, connected the source and got nothing back. The volume that does arrive is concentrated to an extreme degree:

We are flagging this because it should temper how you read every other percentage in this article. One merchant supplies three quarters of the material, so the language mix in particular is really that merchant’s customer base: 63.3% of all Google Shopping reviews we hold are Dutch, 28.9% are English, then 561 Polish, 308 German, 160 Italian and a tail of Czech, Danish and others.

If you run one shop in one country, this does not affect you. If you run the same brand across several markets, it does: seller ratings arrive in whatever language the buyer answered the survey in, so a widget on your German storefront will happily display Dutch text unless you filter by language. That is a setting to check on day one, not after launch.

What to do with the 2,523 ratings that say nothing

15.6% of the Google Shopping reviews we hold, 2,523 of 16,158, are a star rating and nothing else. The buyer answered the survey with a score and skipped the comment box. That share is lower than the 21.5% we see on Google Business reviews, but it matters more here, because the ones that do have text are three times shorter to begin with.

There are two defensible ways to handle them and one bad one. You can count them, which is what a rating summary does: the average and the total both include silent ratings, and that is honest, because a silent five star is still a customer who got their order. You can hide them, which is what a card layout should do, so the widget shows only ratings that carry a sentence. What you should not do is show them as empty cards with a name, a date and five stars and no words. That layout is the single most common reason a review section looks padded rather than trustworthy, and with this source it would fill roughly one card in six.

The same logic decides the placement question from the previous section. A summary can absorb a feed that is mostly numbers. A quote wall cannot, and no amount of styling fixes it.

The source is alive, but half the connections are not

The most recent Google Shopping review in our database is dated 8 September 2026, one day before this measurement. The source works. Whether it works for a given account is a different question: of the 12 accounts holding reviews, 7 have received one within the last 12 months and 5, 41.7%, have nothing newer than a year.

Volume by year tells the same story from another angle:

The peak is 2023, and 2025 is down to a fifth of it. Part of that is accounts going quiet, part is that the post-purchase survey only appears for buyers who opt in and the opt-in rate has never been high. Either way, plan for a seller rating that grows slowly. If you need review volume quickly, this is not the source that delivers it, and the wider picture of which connections still produce is in our audit of review sources that still work.

Getting them onto the page

1. Connect and confirm the count is not zero. Three of our fifteen connected accounts sit at zero and look perfectly connected in the panel. Check the number, not the green tick.
2. Filter by language if you sell in more than one. Seller ratings come in the buyer’s language, and the mix is not yours to choose.
3. Use a summary, not a card wall. With an average of 63 characters of text, a rating summary with the count carries more weight than sparse quote cards.
4. Place it at shop level. Header, checkout, footer, shipping page. Not under a product, where it reads as a claim about that product.
5. Decide about star markup deliberately. Marking up shop-level ratings as if they described a product is the fastest route to a manual action, which we cover in our guide to review schema and stars in Google.

Where we fit

We treat Google Shopping as a shop-level source and default it to a summary layout, which is the correct reading of a feed where nothing is attached to a product. You can filter it by language, combine it with Google Business reviews in one widget so the page is not built on a single survey, and see the real count before committing to a plan. The setup instructions and the embed code are on our Google Shopping reviews widget page, plans are on the pricing page, and you can create an account and connect it to see what your own survey has actually collected.

Two caveats we would rather state than have you discover. Google Shopping is a slow source: our whole database holds 16,158 of these against 388,388 Google Business reviews and 209,357 from Allegro sellers, so it works as a supporting signal rather than as the reviews section of a site. And because it is survey-driven, a merchant who stops the survey stops the feed, which is why five of our twelve accounts show nothing newer than a year old. Connect it alongside another source, not instead of one, and the marketplace reviews guide covers which pairings work.