Why Star Ratings Alone Will Never Win You a Customer

A star rating gets you onto the shortlist. It never closes the decision.

A rating answers: “is this place broadly safe to try?” A testimonial answers: “will this work for me?”

Only the second question was ever standing between a nervous stranger and a booking. The first one they had already assumed the answer to.

This is why a salon with 4.6 and a wall of real, recent customer voices routinely beats a salon with 4.9 and nothing else. The 4.9 got them onto the list. It did not get them the appointment.

What a rating is actually for

It is a filter, and a good one. It is boring, quantitative, and it works at scale: someone comparing three salons on a map at 9pm uses the numbers to eliminate, fast.

Below about four stars, you are gone. Above it, you are through — and that is the entire function. It has now done everything it can do.

Because what a rating is, is an average of other people’s experiences, in general. And the person deciding does not have a general problem. They have a specific one.

The hesitation is always specific

Nobody hesitates over a haircut because they are unsure whether the salon is broadly competent. They hesitate because:

It’ll be too short. She won’t listen and I’ll come out with something I didn’t ask for. It’ll cost more than the website says. He’ll find three other things wrong with the car. It’ll hurt, and I’ll be judged for not coming sooner.

A 4.8 average does not touch a single one of those. It cannot. It is a number about a hundred other people, and their fear is about them, right now, and it is vivid.

Ninety people saying “great” does not neutralise one clear mental image of walking out with hair you hate for six weeks. One woman saying she had exactly that fear, and it did not happen, does.

What a rating cannot show

Three things, all of them decisive, and none of them expressible as a number:

What you are like to deal with. Whether you listen. Whether you make someone feel stupid for arriving with a photo off Instagram. Whether you talk them into something expensive. A star rating is silent on all of it, and for a service that is most of what the customer is buying.

Whether it is still true. A 4.7 accumulated over four years tells you nothing about this year. The good colourist may have left; the owner may have sold up. Everyone who has ever read reviews knows this in their bones, which is why the first thing a careful person does is sort for the most recent ones. They trust what happened last month over an average built up in a busier, better era. People know averages go stale, and they compensate by looking for something recent.

Who it is for. “Great salon!” from a hundred strangers does not tell a woman with very fine hair whether her problem is solved here. A customer saying “I’ve got really fine hair and every place makes it look flat, and — I don’t know what she did” tells her exactly that, and tells nobody else, which is why it converts the person it is for.

The perfect score problem

Here is something the numbers game hides: a flawless average can read as a warning.

Anyone who has bought online knows a wall of nothing but five stars is often bought, seeded by staff, or quietly filtered to bury the bad ones. So a careful buyer discounts it. A 4.9 with three hundred reviews and not one grumble is, oddly, less believable than a 4.6 with a few honest “waited a bit longer than I’d have liked, but worth it” in the mix.

The point is not to chase a lower score. It is that believability, not perfection, is what moves someone — and believability comes from texture, from the odd rough edge that proves a real person typed it. A number cannot carry texture. A testimonial that keeps its stumbles can.

Picture a nervous first-time patient

Take a worked example. Imagine a dental practice with a 4.8 and forty reviews, all some version of “friendly staff, clean, professional.” Genuinely good. And a woman who has not sat in a dentist’s chair for nine years — because the last one made her feel ashamed — is reading it at eleven at night.

The 4.8 does nothing for her. “Professional” is not her fear. Her fear is the lecture: the tut, the “you’ve left this far too long,” the flush of being judged for how long she stayed away.

Now imagine one short clip on that same page: a woman roughly her age saying, “I hadn’t been in years and I was dreading the telling-off, and — there just wasn’t one. He just sorted it.” That is the whole booking. Not because it flatters harder than the average, but because it is aimed. It names her exact hesitation and says it did not happen. The rating was talking about everyone. This is talking to her.

The rating is not yours anyway

Worth remembering when deciding where to put your effort.

Your reviews live on a platform that owns them, ranks them, can remove them, and can change how they are displayed tomorrow without telling you. You are a tenant. And you may never pay for one — Google prohibits incentivised reviews outright, so your only lever is asking politely and hoping.

Your testimonials are yours. You recorded them, you hold the consent, you published them, and a testimonial can be tagged — she gets a notification, she comments, and three hundred people who have never heard of you see a woman they know saying you are good. A review sits on a map, waiting to be looked up by someone already searching for you.

Ratings are found. Testimonials go and find people.

But my rating is only 4.3 — am I sunk?

No. Above the roughly-four-star line the filter has already passed you: a 4.3 and a 4.8 both clear it, and the reader moves on to the real question, which the number never answered for either shop. A 4.3 only hurts if it is your only asset. Pair it with recent, specific voices — a real person from this month naming a fear and saying it did not come true — and the gap to the 4.8 salon with nothing else all but closes. What you cannot afford is to sit on the rating and call it marketing.

So do both, and know what each is for

Ask for reviews, never pay for them, and keep the average healthy — it is the filter that gets you considered at all.

Then win the actual decision with something a rating structurally cannot provide: a real person, from this month, naming the fear the reader is currently having, and saying it did not happen.

And do not polish her. The stumble, the “ehm”, the abandoned sentence is what separates her from the star rating — it is proof that a specific human being said this, unprompted. Tidy it and you have produced something as generic as the number you were trying to improve on. Her words go out exactly as she said them; a testimonial that reads better than the customer speaks is a fake one.

Ask the question the rating cannot answer

Your average is doing its job. Leave it alone.

Ask the next customer: “What were you worried about before you came?” — and put her answer where the stranger reading your 4.7 can see it.

That is the sentence that closes what the number only opened. Why it works is set out in how social proof removes a first-timer’s fear.