Most owners treat reviews as weather. Something that happens to you, mostly noticed when it’s bad, occasionally producing a bad afternoon and a defensive reply you regret.
That framing costs money, because reviews are the one marketing asset that keeps selling after you stop paying for it. A rating you earned in March is still deciding coin flips in November. This is the business case: what a star is actually worth in revenue, where ratings feed into whether anyone finds you at all, and what to do about a review that’s unfair. If you want the broader plan reviews sit inside, that’s restaurant marketing for delivery orders.
The Bottom Line
- Research on Yelp ratings has repeatedly found that a one-star increase is associated with roughly a 5% to 9% rise in restaurant revenue.
- About 84% of people trust online reviews as much as a personal recommendation, and 94% say a negative review has convinced them to avoid a business.
- People read reviews for restaurants and cafés more than for any other type of business, so your rating carries more weight in this industry than in most.
- Reviews are also a ranking input: the local pack takes around 60% of clicks on local searches, and ratings help decide who’s in it.
Lower your delivery costs by 23%
How we reduce costs:
- No delivery vehicle expenses
- Optimized local routes
- Pay-per-delivery model
- Average 23% delivery cost reduction
What a star rating is worth in restaurant revenue
The clearest number in this whole subject comes from work on Yelp ratings, which has consistently found that a one-star increase corresponds to something like a 5% to 9% increase in restaurant revenue (EHL Hospitality Insights). On a restaurant doing a million a year, moving from 3.8 to 4.8 is not a reputation project. It’s a $50,000 to $90,000 project.
Toast’s analysis of ratings data adds the mechanism: restaurants with higher Yelp scores sell tables meaningfully more often during peak periods, the point at which demand is contested and a rating is the tiebreaker (Toast). Review volume, not just the average, also correlates with profitability. Ten reviews at 4.8 does less than two hundred at 4.5, because a small sample reads as untested.
That’s the part worth internalizing: you’re not managing a vanity score. You’re managing the number that breaks ties in every moment of contested demand you have.
Metrobi is transforming catering deliveries
Specialized solutions for catering businesses:
- Catering-trained drivers
- Proper handling equipment
- Peak day delivery support
- 23% average cost reduction
How reviews decide whether anyone finds you
Before a review persuades anybody, it affects whether they see you at all. Around 80% of diners start looking for somewhere to eat on Google, and the three-business local pack at the top of those results takes roughly 60% of the clicks (Chowly). Rating and review volume are inputs into who occupies those three slots.
The same logic runs inside delivery apps, where sorting and ranking lean on ratings, delivery times and order volume. A 4.2 next to three 4.7s isn’t competing on merit; it’s competing from further down the list.
Which gives reviews a double effect that most other tactics don’t have. Better reviews mean more people see you, and a higher proportion of the people who see you choose you. Paid advertising buys the first half only, and stops the day you stop paying.
Where your reviews live, and which ones matter
Not all review platforms are doing the same job, and spreading equal effort across them is a waste.
| Platform | What it drives | How much it matters |
|---|---|---|
| Local search visibility, the local pack, map results | Highest for almost every restaurant | |
| Delivery apps | Ranking and conversion inside the app | High if delivery is a real channel for you |
| Yelp | Discovery in some markets, strong revenue correlation in the research | Moderate to high, varies sharply by city |
| TripAdvisor | Visitors and tourists | High only if you’re in a tourist area |
| Community and older local audiences | Low to moderate |
For most independents the honest ranking is Google first, then whichever delivery apps you actually get volume from, then everything else. If you’re in a tourist market, TripAdvisor jumps up. If you’re in a city where Yelp is still the habit, Yelp does.
The practical implication: send every review request to one place. Split requests across four platforms and you get four thin review counts instead of one credible one.
Why reviews matter more for delivery orders
When someone eats in your room, the room does the persuading. When someone orders delivery, they’re choosing from a list, and the only things on that list are a photo, a price, a delivery estimate and a rating.
That makes your rating a much larger share of the decision than it is for dine-in. It’s also a rating you have less control over, because delivery introduces failure modes that have nothing to do with your food: a cold arrival, a missing side, a long wait. Those show up in your score anyway.
What helps is the stuff that reduces the failures rather than the stuff that manages the reviews: honest prep times, packaging that holds heat, a check on every bag before it leaves. Packaging in particular does double duty here, since it’s both the thing that protects the food and the thing the customer photographs. That side of it is covered in our guide to restaurant branding.
How to get more reviews without annoying anyone
Volume is mostly a function of asking. The restaurants with hundreds of reviews aren’t better liked than the ones with thirty; they have a habit and the others don’t.
One habit, one platform, one moment:
- A card in the delivery bag or a line on the receipt with a code that opens your review page directly.
- One line in your post-order email, at the point the customer has actually eaten.
- Staff asking, at the end of a visit that clearly went well.
Two rules. Don’t reward reviews. Review platforms penalize paid or incentivized reviews, and getting caught is worse than having few reviews. And make the path one tap; every extra step loses most of the people who meant to.
The tool you use for the code matters more than it looks, mainly because some generators stop working if the vendor changes plans and takes your printed cards with them. Uniqode’s guide to selecting a QR code generator walks through the criteria that decide this: editable codes, tracking, and the static-versus-dynamic distinction. Static codes are the trap, because they cannot be tracked or repointed once printed.
How to respond to a bad restaurant review
Responding is not damage control for the person who complained. It’s a message to the next fifty people who read it, who will judge you far more on the reply than on the complaint.
What works:
- Reply to everything, good and bad. Businesses that respond across the board tend to accumulate more reviews and better ratings.
- Answer quickly and shortly. Two or three sentences. Long replies read as arguing.
- Name the specific thing. “The pad thai should not have arrived cold” shows you read it. “We’re sorry you had a suboptimal experience” shows you didn’t.
- Say what changed, if anything did. This is the sentence that convinces readers.
- Move the rest offline. Offer a direct contact and stop there.
What doesn’t: disputing the facts, explaining how busy you were, mentioning that they came on a Saturday, or implying the reviewer is lying. Even when you’re right, you lose. The audience is strangers, and strangers side with the customer.
One negative review among many is also close to harmless, and can help: an unbroken run of five stars reads as fake. It’s the pattern that matters. If three reviews in a month mention the same slow delivery window, that’s not a reputation problem, it’s an operations problem sending you a free report.
Reputation as the cheapest marketing you have
Set reviews against the alternatives and the cost comparison is lopsided. Ads stop the day the budget stops. Commission recurs on every order. A rating, once earned, works on every future search, every app listing, every coin flip, at no ongoing cost.
That’s why review work belongs near the top of any short list of things to do, alongside the other high-payoff basics in our restaurant marketing ideas. It pairs particularly well with social, since the customer who photographs their food is the same one who’ll rate it. See social media marketing for restaurants for turning that into content you can reuse. And a solid rating makes premium mechanics viable, because people buy gift cards for restaurants that look like a safe bet, as covered in the benefits of a restaurant gift card program.
Reputation management is also one of the easier things to hand off, because the useful part is a repeatable habit rather than a creative judgment. If you’re weighing that up, how to choose a restaurant marketing agency covers what it costs and what to ask.
Frequently asked questions
How much do restaurant reviews affect revenue?
Research on Yelp ratings has repeatedly associated a one-star increase with roughly a 5% to 9% increase in restaurant revenue. The effect shows up most clearly during peak demand, when a rating is what breaks the tie between comparable options.
Which review site matters most for restaurants?
Google, for almost every restaurant, because it feeds the local pack and map results where most restaurant searches start. Delivery app ratings come next if delivery is a real channel. Yelp and TripAdvisor matter sharply more in some markets than others.
Should I respond to negative restaurant reviews?
Yes, briefly and quickly, and write the reply for the strangers reading it rather than for the reviewer. Name the specific problem, say what you’ve changed, and offer to continue offline. Never dispute the facts in public.
How many reviews does a restaurant need?
Enough that the average looks tested: a few dozen at minimum, and ideally a steady flow rather than a burst. Review velocity matters as well as volume: a hundred reviews from two years ago reads as a restaurant that may have changed hands.
