Industries
Restaurant reviews widget: Google and Tripadvisor scores compared
A restaurant reviews widget has one job on a menu or booking page: to settle the question the visitor is holding open in another tab before they pick up the phone. Pulled from our production database on 26 September 2026, the 32 food and drink businesses we host hold 10,518 Google reviews between them and average 4.45 stars, against 4.71 across all 391,564 Google reviews in the system. Kitchens are marked down harder than any other trade we serve, and every design decision below follows from that single gap. Here are the figures, then the four choices they settle.
Diners rate harder, and the gap is measurable
We identified food and drink accounts by reading the business name on the connected Google profile, which gives 35 venues: pizzerias, coffee roasteries, a Bäckerei, a catering firm, burger bars, sushi counters and full-service restaurants across Europe, the Americas and Asia. Thirty-two of them hold Google reviews, 10,518 records running from November 2010 to 24 September 2026. The middle account holds 195 of them, the smallest one, the largest 1,762.
| Google ratings, 26 Sep 2026 | Our food and drink accounts | Every business we host |
|---|---|---|
| Reviews counted | 10,518 | 391,564 |
| Five stars | 69.0% | 85.9% |
| Four stars | 18.3% | 7.4% |
| Three stars | 5.9% | 2.1% |
| One or two stars | 6.8% | 4.7% |
| Average rating | 4.45 | 4.71 |
Seventeen percentage points of five-star share is not a rounding difference, and it is not one bad restaurant dragging an average. Of the 27 food accounts holding at least 50 Google reviews, the middle one sits at a 72.2% five-star share and 19 of the 27 fall below 80%. The spread runs from 44% to a single flawless outlier. For comparison, our wider look at how star ratings really distribute found the whole customer base clustered far higher. If your own dashboard reads 4.4, you are an average restaurant, not a struggling one.
The same platform rates a restaurant and a kayak tour differently
This is the figure we did not expect. We hold 26 Tripadvisor listings, and the listing URL states
what kind of venue it is: 17 are Attraction_Review, five are Restaurant_Review, three are
Hotel_Review and one points at a member profile. Split our Tripadvisor reviews by that type and the
platform stops behaving like one platform.
| Tripadvisor listing type, 26 Sep 2026 | Accounts | Reviews held | Five stars | One star | Average rating |
|---|---|---|---|---|---|
| Restaurants | 5 | 1,168 | 47.9% | 5.1% | 4.14 |
| Hotels | 3 | 119 | 93.3% | 0.0% | 4.91 |
| Attractions and tours | 17 | 11,615 | 95.3% | 0.5% | 4.92 |
Same importer, same refresh cycle, same star mapping. A walking tour on Tripadvisor carries a 95% five-star share; a restaurant on Tripadvisor carries 48% and is ten times more likely to be sitting on a one-star. Any advice that treats a Tripadvisor rating as a single benchmark is averaging two different audiences, and the technical detail behind those listings is in our guide to adding Tripadvisor reviews to a website.
Three restaurants that run everything at once
Three of our accounts connect Google and Tripadvisor together, and two of the three add Facebook and Yelp on top. Because it is the same kitchen, the same period and the same customers, the ranking below is the cleanest read we have on what each platform does to a restaurant’s score. It is ordered by how flattering each platform is, sternest first.
| Same three restaurants, every platform they run | Reviews supplied | Five stars |
|---|---|---|
| Tripadvisor | 124 | 50.8% |
| Yelp | 97 | 55.7% |
| 650 | 68.8% | |
| 348 | 94.5% |
Two things fall out. Google supplies five times the volume of Tripadvisor for these venues, so a page built on Tripadvisor alone is a page built on a fifth of the available evidence. And the spread from 50.8% to 94.5% means the platform badge on each card is doing real work: without it, a visitor reads a suspiciously mixed set instead of four different rooms asking four different questions.
Facebook is the extreme case and worth understanding before you embed it. Facebook dropped star ratings for a yes or no recommendation, so across all 49,193 Facebook reviews we hold, exactly zero are four, three or two stars. It is 96.3% five and 3.6% one, and nothing else exists. That flat 5.0 next to a 4.4 from Google is not a better restaurant, it is a different question, and the card has to say so.
What restaurants actually connect, and what quietly fails
Of our 35 food and drink accounts, 22 run Google alone, 12 run Google plus Facebook and one runs Google plus Yelp. Not one of the 35 has a Tripadvisor connector. The restaurants in our Tripadvisor set are a separate group of businesses that mostly do not run Google with us, so in practice the two crowds barely overlap.
The plumbing explains part of that. Of the five Tripadvisor restaurant listings we hold, only one is
crawling cleanly right now: three return job_not_found and one is an invalid URL, usually a listing
that was renamed or moved after it was pasted in. Yelp is thinner than its reputation for restaurants
too. We hold 93 Yelp connectors, but reading the URLs they are mostly lawyers, movers, clinics and
car services; only a handful are food, and across the food accounts Yelp has supplied 13 reviews in
total. Zomato exists in our data as exactly one connector, 191 reviews, averaging 4.25. Before you
plan a widget around a platform, it is worth checking which review sources still
work rather than which ones you remember.
Four decisions the numbers settle
Start with Google and put it next to the booking action. It supplies the volume, and for 22 of our 35 venues it is the only thing connected. Our Google reviews widget runs from the public profile, refreshes every 24 hours on every plan including the free one, and belongs beside the reservation button rather than in the footer.
Add Tripadvisor second if you have a listing, and let it be the stern one. A restaurant page carrying 4.7 from Google and 4.1 from Tripadvisor reads as collected. A page carrying 4.9 everywhere reads as chosen. Our Tripadvisor reviews widget takes the public listing URL, so check the URL still resolves before you paste it, given that four of our five restaurant listings broke exactly there.
Keep the platform badge on every card. With a 50.8% to 94.5% spread across platforms, the badge is the difference between a mixed set that looks verified and one that looks edited.
Show the real distribution, not a hand-picked five. A 4.45 average with visible four-star cards is the ordinary shape of a working restaurant, and the industry benchmark puts it in context for anyone who wants to argue about it. You can open an account and have Google cards live on the menu page today.
On plans, so nothing surprises you: the free tier carries Google and Facebook only, which covers 34 of our 35 food accounts as they stand, and Tripadvisor needs a paid plan. Paid plans open with a 7-day trial and the full ladder sits on the pricing page.
The caveats, in the order they matter. This is 35 venues, not a national survey, and they skew independent rather than chain, so read the direction of the gap rather than the decimal. We found them by reading Google profile names, which misses any restaurant whose name contains no food word, so the true count in our base is higher than 35. The Tripadvisor restaurant sample is five accounts and one of them supplies 808 of the 1,168 reviews. And none of it changes a booking until the snippet is actually on the page, next to the table you want filled.
FAQ
Which review source should a restaurant put on its own site first?
Google, by a wide margin, and the reason is volume rather than loyalty. Counted on 26 September 2026, our 32 food and drink accounts hold 10,518 Google reviews with a middle account at 195, while the five restaurants running a Tripadvisor listing with us hold 1,168 between them. Of the three restaurants that connect both, Google supplied 650 records against 124 from Tripadvisor. Add the second platform once the first is live, not instead of it.
Why is my restaurant scored lower than other local businesses?
Because kitchens are. Across our food and drink accounts the Google average is 4.45 stars with 69.0% of ratings at five, while across all 391,564 Google reviews we hold the average is 4.71 with 85.9% at five. Of the 27 food accounts holding at least 50 reviews, 19 sit under an 80% five-star share. A 4.4 on a menu page is an ordinary restaurant score, not a warning sign, and the widget should not be tuned to hide it.
Is Tripadvisor worth connecting for a restaurant?
Only as a second source, and expect sterner numbers. Within our own Tripadvisor data the listing type decides the rating: 17 attraction listings average 4.92 stars with 95.3% five-star, while the five restaurant listings average 4.14 with 47.9% five-star and 5.1% at one star. It is the same platform and the same importer, so the gap is diners, not scraping. The upside is credibility, since a page mixing 4.9 and 4.1 reads as collected rather than curated.
Why does my Facebook rating look perfect next to Google?
Because Facebook has no middle. Facebook replaced star ratings with a yes or no recommendation, so across the 49,193 Facebook reviews we hold exactly 0.0% are four, three or two stars: 96.3% land at five and 3.6% at one. Put Facebook cards beside Google cards without a platform mark and the mix looks edited. Showing the source badge on every card is what makes the flat 5.0 read as a different question rather than a better restaurant.