In Defence of the Dish: Why Star Ratings Should Stop Running Discovery
Star ratings have been the centre of restaurant discovery for a decade. They're the wrong metric — and it's time to replace them with facts.
A restaurant is more than its star rating
The entire restaurant discovery industry, for the better part of a decade, has been organised around a single number: the star rating. It is the default currency of trust. It is also, quietly, one of the least useful pieces of information a diner can be given.
A star rating is a backward-looking average, diluted by strangers, unweighted for relevance, and stripped of the context that actually makes a restaurant worth visiting. It tells you that 243 people, on average, had a decent time. It does not tell you what's on the menu, what it costs, whether the dish you want is any good, or whether the place is a fifteen-minute drive or a ninety-minute one.
We have built an entire discovery layer around the wrong metric. And the cost of that mistake is visible everywhere — in restaurants gamesmanship of reviews, in diners' distrust of suspiciously high ratings, in the weary ritual of reading a dozen reviews to extract a single useful fact.
It is time to retire the star rating as the centre of discovery, and replace it with something more honest.
The argument
A star rating answers "was this probably fine?" A diner actually asks "does this restaurant serve the dish I want, at a price I accept, near here?" The industry optimised for the wrong question.
What a star rating can't tell you
Consider what you actually need to decide on a restaurant.
You need to know what they serve — and a rating doesn't. You need to know what it costs — and a rating doesn't. You need to know where it is — and a rating doesn't. You need to know whether the specific dish you're craving is any good — and a rating gives you a weighted average of every dish on the menu, which is meaningless.
A 4.6-star restaurant with an average burger and an exceptional dessert will look identical, in rating terms, to a 4.6-star restaurant with an exceptional burger and an average dessert. The diner who wanted a burger will go to the wrong one.
1 number
The star rating
Answers almost no useful question
0 context
No dish, price, or proximity
The questions diners actually ask
The distortions ratings create
Ratings don't merely fail to inform. They actively distort the market.
Restaurants chase the metric, not the meal. Once a rating becomes the primary discovery signal, restaurants optimise for ratings — sometimes legitimately, sometimes not. The energy that should go into the menu goes into the review funnel. The incentive is to please reviewers, not necessarily diners.
Diners trust ratings less and less. It is now common knowledge that ratings are gamed, that aggrieved reviewers are louder than satisfied ones, and that a 4.8 and a 4.3 might be functionally identical. The number has lost meaning precisely because it was asked to carry meaning it couldn't support.
Independent restaurants are disadvantaged. A new independent has fewer reviews, more volatile averages, and no marketing budget to solicit ratings. A chain starts with volume on its side. A rating-centred system systematically privileges the players who already had the advantage.
What should replace it
The honest replacement for the star rating is not a better rating. It is data.
The questions a diner actually asks — what do they serve, what does it cost, where are they, is the dish I want good, are they open — are all answerable with structured information, not opinion aggregates. A dish-first platform that indexes the menu, surfaces the price, respects the geography, and tags the dietary need is answering every question a star rating fails to, in real time, without the dilution of a thousand strangers' moods.
Most-searched dishes this month
Live search trends — renders automatically as platform data grows
Where ratings still belong
Ratings are not useless. They are a trailing confirmation, not a leading decision tool. Once you've found the restaurant that serves your dish at your price in your area, a rating is a reasonable final sanity check. But it should be the end of the decision, not the beginning. We have been using it as the beginning for fifteen years, and the result is a discovery experience diners tolerate rather than trust.
The principle
Discovery should start with what a restaurant is — its menu, its prices, its location — before it considers what people thought of it. The facts precede the opinions. That inversion is the entire premise of Dine South Africa: structure first, sentiment second.
A restaurant's menu, priced and tagged and searchable, is more useful to a hungry diner than a 4.7-star average could ever be. The industry has spent a decade refining the wrong number. It's time to stop averaging opinions and start surfacing facts.
The verdict
Retire the star rating as the discovery-centre it was never suited to be. Replace it with the structured menu, price, and location data that answers the questions diners actually ask. Let ratings be the confirmation. Let the facts be the discovery.
— The Dine South Africa Insights Team
Related: Why Choosing a Restaurant is Harder Than It Should Be · If Your Restaurant Menu Isn't Searchable, It Isn't Working
Dine South Africa Insights
Opinion
