Download Google Reviews for a Restaurant — Food, Service & Atmosphere Scores to Excel (2026)

To download Google reviews for a restaurant, copy the restaurant's Google Maps link, paste it into a Google reviews exporter and download Excel or CSV. You get every public review as a row: star rating, text, date, the owner's reply, photos, and on most restaurant reviews the separate Food, Service and Atmosphere scores, meal type and price per person that Google asks diners for. The first 100 reviews need no signup, and you do not need to own the listing.

Restaurants are among the most common businesses behind the Google reviews exports we run, so this guide is built from those files: what's in them, which columns restaurant owners actually use, and the formulas that turn a few thousand reviews into a menu and service report.

Why restaurants pull their reviews into a spreadsheet

Google Maps shows reviews a few at a time, sorted by "most relevant". That's fine for a diner choosing where to eat and useless for an owner trying to understand a pattern. A spreadsheet answers questions the app can't:

• Which dishes come up most often, and are they praised or complained about?
• Is the problem the food, the service or the room?
• Did the rating drop after a menu change, a new chef or a busy summer?
• Which staff members do guests mention by name?
• How fast, and how often, does the owner reply to bad reviews?

The stakes are real. A Harvard Business School study of restaurant reviews found that a one-star increase in rating led to a 5–9% increase in revenue for independent restaurants. Knowing why the stars move is worth an afternoon in a spreadsheet.

Getting every review out of Google Maps

1. Find the restaurant on Google Maps and open its listing.
2. Tap Share and choose Copy link. On desktop, the long google.com/maps/place/… address from the browser bar works too.
3. Paste it into the Google reviews exporter. You see the number of reviews before anything downloads.
4. Download Excel to open straight in Excel, or CSV for Google Sheets.

No Google login, no Business Profile access and no Places API key are involved. The official Places API returns only 5 reviews per place, which is why owners who try it first end up here. Longer review histories need a one-time payment, never a subscription; see the pricing page.

The columns a restaurant owner actually uses

A Google reviews file has more than 25 columns. These are the ones that matter for a restaurant:

• rating — the overall 1–5 stars.
• text — the review as written, in its original language.
• translated_text — Google's English translation when the review is in another language.
• created_at — when it was posted, for trends by month or season.
• owner_reply and owner_reply_at — your reply and when you sent it.
• replies — 1 if the owner replied, 0 if not, so =AVERAGE() on it is your reply rate.
• photo_count and photo_urls — the pictures guests attached, often of the food.
• attributes — the restaurant-specific details described next.
• author_is_local_guide and author_review_count — whether the reviewer is an experienced Google reviewer or a one-off.

Food, Service and Atmosphere: the scores hiding in the attributes column

When someone reviews a restaurant on Google, Maps invites them to rate the food, the service and the atmosphere separately, and to add details like meal type and price per person. Those answers export in the attributes column, looking like this:

{"label":"Meal type","value":"Dinner"}; {"label":"Price per person","value":"€20–30"}; {"label":"Food","value":5}; {"label":"Service","value":1}; {"label":"Atmosphere","value":2}

That one row tells a story the overall stars hide: the food was excellent, the service was terrible. Across a few hundred reviews, these sub-scores show which part of the restaurant is costing you stars.

To split them into their own columns, assuming attributes is in column T:

• Google Sheets: =IFERROR(VALUE(REGEXEXTRACT(T2, """Food"",""value"":(\d)")), ""). Swap Food for Service or Atmosphere in the next two columns. See Google's REGEXEXTRACT reference.
• Excel (Microsoft 365): =IFERROR(VALUE(LEFT(TEXTAFTER(T2, """Food"",""value"":"), 1)), ""), using Microsoft's TEXTAFTER function.

Average each new column and you have three numbers instead of one. Filter to reviews where Food is 5 but Service is 1 or 2, and you are reading exactly the complaints a front-of-house fix would solve.

The same column carries Meal type (breakfast, lunch, dinner), Price per person, Group size and Reservation. Pivot ratings by meal type and you may find that lunch service is fine and dinner is where things fall apart.

Finding dishes and staff names in the text

Guests name what they ate and who served them. To count mentions, put a list of dishes in one column and use:

• How often a dish comes up: =COUNTIF(C:C, "*risotto*")
• Its average rating: =AVERAGEIF(C:C, "*risotto*", L:L), where L is the rating column.

A dish that appears in many reviews with a low average is a menu problem. A dish with high mentions and a high average belongs on your social posts. Run the same formulas over the translated_text column so foreign-language reviews count too.

Staff names work the same way. In one Italian restaurant's export of 9,334 reviews, a single waiter was thanked by name in 454 of them, mostly by Arabic-speaking guests. No amount of scrolling on Maps would have surfaced that. It's a training example and a staff reward in one formula.

Tracking your rating month by month

The overall rating on Maps is an average of every review the restaurant has ever had, so it moves slowly and hides recent changes. Your file shows what is happening now:

1. Add a column with the month of each review: =LEFT(F2, 7) on created_at gives 2026-09.
2. Build a pivot table with the month as rows and the average of rating as values. Add the Food, Service and Atmosphere columns you extracted.
3. Chart it. A line that dips in July and August and recovers in September is a summer staffing problem, not a kitchen problem.

Mark the dates of real changes on the chart: a new menu, a new chef, a renovation, a price rise. If the Food score moves the month after a menu change, you have evidence instead of opinions. Download again every few months and stack the new file under the old one, removing duplicates on id, to keep the history growing.

Reviews in other languages

Restaurants in tourist areas collect reviews in a dozen languages. In the restaurant reviews we exported over the last 90 days, about 45% came with a Google English translation, meaning they were written in another language. The original text stays exactly as written, and the translation sits next to it, so you can read everything in English while keeping the original for quotes and replies.

Auditing your replies

Google recommends that businesses reply to reviews, and diners notice when owners do. With the file open:

• Reply rate: average the replies column.
• Reply rate on bad reviews: =AVERAGEIFS(E:E, L:L, "<=2") (E = replies, L = rating). This is the number that matters most.
• Reply speed: subtract created_at from owner_reply_at.
• Unanswered 1-star reviews: filter rating to 1 and replies to 0. That's your to-do list.

Comparing your restaurant with the ones nearby

Your competitors' reviews are public too. Export the three to five restaurants your guests compare you with, add a restaurant column to each file and stack them. Then compare:

• Average Food, Service and Atmosphere scores, side by side.
• Price per person against rating: who delivers more for the money.
• The complaints that appear at every restaurant (parking, noise, waiting) against the ones unique to you.
• Their best-loved dishes, which tell you what the local market rewards.

With a pass you can paste up to 20 Google Maps links at once, one per line, and download every restaurant in one batch. Our Google reviews competitor analysis guide goes deeper on the comparison itself.

What 12,755 restaurant reviews taught us

This guide comes from operating the exporter. Over the 90 days to 23 October 2026 we ran 133 Google reviews exports, and the restaurant listings among them held 12,755 reviews. What we saw:

• 79% of restaurant reviews carried the sub-scores for food, service or atmosphere in the attributes column. Most owners we talk to had never seen them in a spreadsheet.
• 55% had an owner reply, against 27% across all the Google reviews we exported. Restaurants reply far more than most businesses, which makes an unanswered bad review stand out.
• About a third of Google reviews have no text, only stars. They still count toward the average, so they are worth keeping in the file.
• Large restaurants are normal. The biggest single restaurant we exported, an Italian pizzeria, had 9,334 reviews: more in English, German, Arabic and Polish than in Italian.

The column names, the 20-link batch and the first 100 reviews with no signup are the live product values. For what may appear in reviews and how they are moderated, Google's Maps user-generated content policy is the authority.

Using the file without upsetting anyone

Reviewer names and profile links are personal data. Use the file to improve the restaurant, quote short extracts without names in internal reports, and never contact reviewers through details found in the export. Replies belong on Google, where every future diner can see them.

Download a restaurant's Google reviews →

Related reading: downloading all Google reviews of a business, Google reviews competitor analysis, why the Places API returns only 5 reviews.