Service retention
Service retention explained: how it is measured, and why two dealerships never mean the same thing by it
Service retention is the share of a dealership's customers who return to that dealership for service inside a defined window, usually measured against the vehicles it sold. It tracks whether the relationship survived the sale, and it is the fixed operations metric most sensitive to how you define it.
Also called customer retention, service loyalty, or owner retention, and reported by manufacturers under their own franchise-specific names. Every one of those definitions is slightly different from yours.
How is service retention calculated?
Service retention = customers with at least one qualifying service visit in the window ÷ eligible customers in the window × 100
Simple arithmetic, three loaded terms. Each has to be written down before the number leaves the room:
- The window. Twelve months from delivery is the most common. Twenty four months is also common and always produces a higher figure, because it gives every customer two chances instead of one.
- The denominator. Vehicles the dealership sold, all vehicles of the franchise in the catchment area, or every unique vehicle that has ever visited the workshop? These are three different businesses being measured.
- What counts as a visit. Any repair order, or a paying visit only? A safety recall the manufacturer summoned the customer for is not evidence of loyalty, and neither is a warranty rectification.
Service retention: a worked example
A dealership delivered 1,400 vehicles last year. Twelve months after each delivery, 812 of those vehicles had come back for at least one paid service visit. Retention is 812 ÷ 1,400 = 58.0 percent.
Now hold the customer behaviour completely still and change only the reporting. Stretch the window to 24 months and a further 140 vehicles appear, giving 68.0 percent. Count any repair order rather than paid visits only, and the recall and warranty visits push it higher again. The workshop did not get better between those three figures. The spreadsheet did.
So a retention number quoted without its window and its denominator is close to worthless, and the figure on the manufacturer's franchise dashboard rarely matches the one out of your own DMS. Both can be correct.
What the current data shows
The Cox Automotive Service Industry Study, surveying 1,974 US vehicle owners aged 18 to 75 between April and May 2025, found that only 54 percent of owners with vehicles two years old or newer returned to the dealership where they bought the car, down from 72 percent in 2023. Across the market, dealerships handled 12 percent fewer service visits than in 2018.
The reason customers gave is the part worth sitting with. Nearly half of owners, 45 percent, reported dissatisfaction with their dealership service experience, primarily citing unexpected costs and poor communication. And dealership repair costs in 2025 averaged 261 dollars against 275 dollars at general repair shops (Cox Automotive, 11 November 2025).
So the dealership was cheaper on average and lost the customer anyway. Price was not the mechanism. Being surprised by a bill, and not understanding what was done or why, was.
The same study found 74 percent of owners who service at a dealership say they are likely to buy their next vehicle there. Retention is not only a fixed operations number, which is why it ends up in front of the dealer principal alongside service absorption.
How service retention gets misread
- Reported as an average. Retention in years one and two behaves nothing like retention in years four to six, when the warranty expires and the independent down the road becomes a live option. The blended figure hides exactly the cohort that is leaking.
- Recall visits counted as loyalty. A manufacturer campaign brings customers in who had already defected. Strip them out or the number flatters the department that lost them.
- Compared against the OEM score. Manufacturer retention reporting uses its own window, its own denominator and often its own catchment definition. Track both, never reconcile them, and never set a target on one using the other's history.
- Measured on repair orders instead of vehicles. Retention on repair order count rewards frequent visits from a shrinking base. A dealership can lose a third of its customers and hold the visit count steady for a year.
- Confused with defection. A customer who has not returned in 14 months has not necessarily gone anywhere. They may drive 4,000 miles a year. Segment by mileage first.
- Treated as a marketing problem. A service reminder brings back a customer who was already willing. It does nothing for one who felt sold to.
- Read without CSI. Retention is lagging by design. CSI and declined work move first, and they tell you what retention will do next year.
Where retention is actually won or lost
Not at the reminder. At the write-up and at the moment work gets recommended.
A customer who does not understand why a repair is needed has two ways to read the recommendation: as expert advice, or as a dealership finding things. They cannot tell which from a phone call quoting a number, and the Cox finding on unexpected costs and communication is that doubt showing up in a survey a year later. The customer who says no, then quietly gets it done elsewhere, has already left. The retention report notices next year.
Which is why a vehicle health check the customer can actually see changes the conversation. When the technician films the worn pad or the leaking seal and the customer watches it on their phone, the recommendation stops being a claim and becomes something they can judge for themselves. VentaVid builds personalized video for service teams for exactly that step, and reports that 85 percent of service videos sent through its platform are viewed within 15 minutes, which is VentaVid's own reported figure rather than an industry benchmark. How it fits a live write-up process is covered in more depth in service video software for automotive.
