The Anxiety of 4.5
A beauty salon. Google rating 4.5. Nearly all five stars. The owner is satisfied.
Until one month, the repeat visit rate drops 15%.
She’s confused. The rating is so high — why aren’t customers coming back?
She opens the dashboard and checks the numbers. New customers haven’t decreased — Google Maps visibility remains stable, and the number of people clicking through hasn’t changed. But the percentage of customers who visited within the last three months and booked again dropped from 42% to 27%. Where did that 15% go?
She asks a few regulars. One says, “I’ve been busy lately.” One leaves her on read. One says outright, “I tried another place — better value.”
But these are anecdotal. What she needs is a structural explanation — not “the personal choices of a few customers” but “is there some systematic reason pushing customers away?”
The answer is right there in her Google reviews. Not in the star count. But in the text beneath the stars that she never properly analyzed.
Averages Deceive You
The 4.8 she sees is the average of all ratings. But averages have a fatal weakness: they flatten all differences into a single number.
Suppose the five dimensional scores are: Technique 5.0, Environment 4.9, Service 4.8, Booking Convenience 4.7, Value Perception 3.2.
Averaged: 4.52. But Google displays the average of overall star ratings — a flood of five-star reviews overshadows the few low scores, and after rounding, it shows 4.5. Still looks fine.
Customers leave five stars because the technique is excellent. Customers don’t return because value perception is low — “It looks great, but it’s expensive. I don’t feel it’s worth it.”
The 4.5 masks the 3.2. “Looks good” masks “not worth the price.”
What’s more dangerous is that this 3.2 never appears in any single review’s star rating. No customer gives one star for “feels a bit expensive.” They give five stars because the technique genuinely is excellent. But within the text of those five-star reviews, if you read carefully, you’ll spot clues: “The results are great, but the price really isn’t cheap.” “The stylist is very professional, just a bit hard on the wallet.” “I’d recommend it to friends, but I’d warn her about the price first.”
These sentences are buried inside five-star reviews, eclipsed by the brilliance of the stars. But they are ticking bombs.
Anscombe’s Quartet
In 1973, British statistician Francis Anscombe published a four-page paper in The American Statistician titled “Graphs in Statistical Analysis.” He did something very simple but very striking. He constructed four datasets, each with: identical means (all 9.0), identical variances, identical correlation coefficients (0.816), and identical regression lines (y = 3 + 0.5x).
Looking at the numbers alone, all four datasets are “identical.”
But when he plotted them — four completely different patterns. The first is a normal linear distribution. The second is a graceful curve with an obviously poor linear fit. The third has a single outlier that completely distorts the entire line. The fourth is nearly vertical — all points clustered together, with a single outlier determining the regression line.
Anscombe’s point: If you only look at summary statistics (mean, standard deviation), you’d think all four datasets tell the same story. But the graphs reveal they tell four completely different stories.
This paper later became a classic in statistics education, changing how countless analysts approach data. Tukey’s Exploratory Data Analysis, published in 1977, took this principle to its extreme — plot first, then examine the numbers.
Your 4.5 rating is Anscombe’s mean. It compresses the story of five dimensions into a single number. Without examining the dimensional distribution, you have no idea where your 3.2 is hiding.
Why Five Stars Are More Dangerous Than One Star
One star forces you to act. Five stars lull you into complacency.
A 4.5-rated salon won’t convene an emergency meeting, won’t review processes, won’t pressure staff. Because “the score is high, no problem.”
But the 3.2 value perception score is quietly eroding your customer base. Customers won’t leave a one-star review saying “not worth the price” — they leave four or five stars saying “the technique is excellent,” and then… never come back.
Customers who don’t return don’t leave negative reviews. They simply vanish. Your 4.5 rating won’t drop — because the only people still leaving reviews are the satisfied ones. Dissatisfied people don’t even bother reviewing anymore.
This is how survivor bias manifests in five-star reviews: high scores mask customer attrition.
During World War II, U.S. Navy analyst Abraham Wald was tasked with studying “how to reinforce bomber armor.” He examined a batch of returning bombers and found bullet holes concentrated on the wings and tail. The intuitive response: reinforce the areas with the most bullet holes.
Wald said no. He said: these are the planes that “came back alive.” The planes hit in critical areas already crashed into the sea. The places with few bullet holes (engines, cockpit) are where armor is truly needed — because planes hit there never made it back.
Your five-star reviews are those “bombers that came back alive.” The customers who didn’t return are the planes that crashed into the sea. They left no bullet holes (negative reviews), but they are the group you most need to pay attention to.
How to Find the Hidden 3.2
You need to break down your overall score into dimensions. Instead of asking “are customers satisfied,” ask “which aspects are customers satisfied with, and which aspects are they dissatisfied with?”
Five dimensions: Product Quality, Service Experience, Environment & Atmosphere, Booking Convenience, Value Perception. Tag each of your reviews one by one — which dimension does this five-star review primarily praise? Which dimension does this four-star review hint at dissatisfaction?
This isn’t difficult, but it’s time-consuming. A hundred reviews, reading each one and classifying it across five dimensions — a trained person needs roughly two hours. If you have three months of reviews (assume 40 per month), that’s six hours.
But the harder part is staying objective. You’re the owner. When you read five-star reviews, your subconscious amplifies positive signals. When you read negative reviews, your subconscious activates defense mechanisms — “this customer was difficult anyway” or “they don’t understand our pricing strategy.” This isn’t unprofessional of you — it’s human confirmation bias. British psychologist Peter Wason demonstrated this in his 1960 experiment: humans are naturally inclined to seek evidence that supports their existing beliefs rather than challenge them.
After tagging a hundred reviews, you’ll see a distribution chart. The lowest dimension in that chart is your 3.2 — even if your overall score is 4.8.
Once you find it, what then? Not immediately slashing prices. Not immediately issuing discount coupons. But understanding: why did customers rate this dimension low? Is the pricing strategy flawed? Is communication insufficient — do customers not understand why your service is worth the price? Or are there too many add-on fees revealed only at the end?
Finding the 3.2 is just the first step. Understanding the cause behind it is what truly transforms your repeat visit rate.
Download the Industry Benchmark Report
How do your five-dimensional scores compare to the industry average? Leave your email to access the industry benchmark report and see whether your “value perception” score is also part of a structural pattern.
Further Reading
← Your reports are beautifully formatted and logically coherent, but they don’t tell you why customers stopped coming
Your reports only look at total scores? That’s just as dangerous as only looking at average ratings. Survivor bias doesn’t just happen with orders — it happens with reviews too.
→ Read
Loading…