A 4.8-star rating with 400 reviews and a $15 average price sounds like a reliable lunch option. Then you go and the food is mediocre, the service is slow, and nothing about the experience corresponds to what that rating implies. As someone who has had this experience enough times to actually investigate what’s going wrong, I learned how to read restaurant ratings in a way that produces much better results. Today I’ll share what I found.
Rating Inflation Is Structural
Most review platforms have a ratings distribution that’s skewed toward the high end. On Yelp, the median restaurant rating in most cities is around 3.7–4.0. On Google, it’s higher. On delivery apps, it’s higher still because the friction of leaving a negative review is low but so is the baseline expectation — people are rating takeout, not fine dining. The result is that a 4.2 on a delivery app represents something different from a 4.2 on Yelp, which represents something different from a 4.2 on a travel review site.
Practically, this means the absolute rating number is less useful than its position relative to comparable restaurants in the same area and category. A pizza place with 4.4 in a market where the average pizza rating is 4.3 is ordinary. A pizza place with 4.4 in a market where the average is 3.8 is actually meaningfully above average. That’s what makes the relative comparison endearing to anyone who uses ratings seriously — the context changes the meaning of the number entirely.
Review Velocity and Recency Matter More Than the Average
A restaurant with 400 reviews accumulated over 8 years has a significantly different signal than a restaurant with 400 reviews accumulated over 18 months. Recent reviews represent the current experience. Reviews from 3 years ago may reflect a kitchen team, owner, or price point that no longer exists. The rating shown is a lifetime average; filtering for recent reviews (most platforms allow this) gives you a more accurate picture of what you’re currently ordering from.
I’m apparently someone who learned this lesson by revisiting a place I’d loved years earlier that now had a 4.4 average — which looked great — until I filtered to the last six months and found mostly 2 and 3-star reviews mentioning the same ownership change.
Photo Volume and Source Are Signals
Restaurants with a high ratio of professional or semi-professional photos to total reviews have usually run a promotion encouraging customers to share photos, or have done a photo session specifically for their listing. Restaurants where most photos are clearly taken by customers on phones at real meals show you the actual product more accurately.
The Text of Reviews Beats the Stars
A restaurant with a 4.2 average that has five recent 3-star reviews mentioning the same specific problem — slow service on weekends, inconsistent portion sizes, a quality drop in a specific dish — gives you more actionable information than the aggregate rating does. The text is where the real information lives. The star is a summary that compresses all of that nuance into a single number calibrated to whatever scale the particular reviewer uses. Probably should have led with this, honestly — the most reliable signal is a restaurant with a moderate rating and a high number of detailed text reviews discussing specific dishes. Detail correlates with authenticity, and authenticity correlates with accuracy.
Leave a Reply