Introduction
Face to face, UK car retail is good at its job. From the moment a customer walks into a showroom the sales experience scores in the high 80s and 90s. The service lane, at the same sites, never gets there. It runs lower at every stage, and the two lowest scores anywhere in the dataset both sit inside it.
Look closer at where those low scores fall and a pattern appears. They cluster at the moments where the customer depends on you to be straight with them and has no way to check.
In a showroom a customer can compare, negotiate and walk away. In the service lane they depend on your diagnosis, your estimate and your explanation. That’s where trust is breaking, and the moments are nameable.
What the data reveals: the lowest scores sit where the customer can’t verify.
Repair work scores 35. Discounts and rebates score 13, the lowest reading in the whole dataset, below price itself. In negative reviews, discounts appear nearly fifty times as often as in praise, diagnosis twenty times, and “didn’t listen to my concern” nineteen times; repair work alone carries a quarter of all negatives. Service runs 47 before arrival, climbs to the 60s when a human appears, then drops to 50 at payment.
That discount number deserves a second look. A concession scoring far below the price itself suggests the discount is being experienced as a reveal (there was a better number all along) rather than as a benefit. What customers reject isn’t the cost. It’s the arrival of a number they didn’t expect, from someone they had to trust.
The reading that follows is editorial, not a cause the data establishes: a showroom customer has options, a service customer has a professional’s word. That dependence is a plausible reason the experience is more fragile here. What the data does establish is where it’s fragile: the diagnosis, the estimate, the discount, the invoice.
Why it matters: the categories falling across the half-year are the trust ones.
Over the six months, the staff-conduct categories held in the high 90s. What fell was satisfaction (down three), communication (down three) and vehicle delivery and return. The platform’s own category tags, measured independently of the language analysis, land in the same place: what breaks is communication and commitment, not competence. Two instruments, one conclusion.
And this is the lane the next decade is leaning on. Separately sourced, and not combined with the review data: an electric vehicle generates around 40% less aftersales revenue over its life than a petrol equivalent; only about a fifth of UK technicians are EV-trained against a projected shortfall of 35,000 by 2030; the weak 2020–21 registration years have hollowed out the used stock that feeds service. Each is a pressure on aftersales in its own right. Service is not a secondary problem that sales performance can offset.
Three ways the business runs differently
Stop explaining the repair. Start confirming the concern. “Not listening to my stated concern” is a communication failure before it’s a technical one. On Monday: every job card opens with the customer’s own words, read back to them.
Stop quoting and invoicing. Start managing the distance between them. Confirm the estimate before work starts, tell the customer before the number changes, and walk the invoice through before payment. On Monday: a count of how often the final figure differs from the quoted one, and by how much.
Stop giving discounts. Start giving the right number first. A concession that reads as a reveal costs more trust than it buys. On Monday: a look at every discount given last week and why it was needed.
Where Reputation comes in
See which moment of the service visit is breaking, at which sites, and whether it’s improving. Reputation maps what customers say to the stages and categories of the service journey, compares them across locations and tracks movement over time, so an aftersales leader can see that estimate complaints fell and diagnosis complaints didn’t, instead of one number that hides both.
The question for your next meeting: at which moment in our service process does the customer stop trusting what we tell them, and would we know?
Three-sentence test. I learned that service breaks at the moments the customer can’t verify, and the data names them. I’ll make the diagnosis, estimate, discount and invoice transparent, with a count of how often the number moves. Reputation could show me which moment is failing, at which sites, and whether the fix is working.




