Dealership Service Retention Is an Information Problem, Not a Price Problem

Abstract

Franchise dealerships are losing service work, and the loss is measured rather than asserted. Dealerships handle 12% fewer service visits than in 2018. Among owners of vehicles two years old or newer, the segment a dealer should hold most easily, the return rate to the selling dealership fell from 72% in 2023 to 54% in 2025. Dealer share of all service visits in the United States fell from 35% in 2021 to 30%, with independent repair at 28% and closing. The industry reads this as a price problem and responds with discounting. The same research refuses that reading: of the 45% of owners dissatisfied with dealership service, the dissatisfaction is attributed to unexpected costs and poor communication, and a customer quoted a figure in advance and charged that figure does not experience it as unexpected, whatever the figure is. Service visits per owner per year rose from 2.3 in 2021 to 2.5 in 2023 while dealer share fell, which establishes that demand did not move but its destination did. This paper argues that service defection is an information failure, draws on the expectation-disconfirmation literature (Hien and Long 2024) and on the customer-defection literature (Piha and Avlonitis 2015) for the mechanism, and sets out what a service department must record for the failure to stop recurring. It states its own limits, including that no causal claim is made from the correlation between communication practice and retention.

What the numbers establish, and what they do not

Three series describe the loss, and they are independent of one another.

Visit volume: dealerships handle 12% fewer service visits than in 2018 (Cox Automotive 2025). Market share: dealer service lanes fell from 35% of all service visits in the United States in 2021 to 30%, while general repair and service stations sit at 28%. Retention of the best-held segment: owners of vehicles two years old or newer returned to the selling dealership at 54% in 2025, down from 72% in 2023.

A fall of eighteen percentage points in two years (Cox Automotive 2025), on the segment still inside or near its warranty period, is not a slow structural drift. It is the segment where the dealer holds the strongest claim, because the vehicle is new, the relationship is recent and the warranty work is theirs by default.

The fourth series is the one that constrains the explanation (Cox Automotive 2023). Service visits per owner per year went from 2.8 in 2018 to 2.3 in 2021 and back to 2.5 in 2023. Demand recovered while the dealer's share of it fell. The work did not disappear. It moved.

That rules out an entire family of explanations. Vehicles have not become more reliable in a way that removes the work, because the work is being done. Owners have not deferred maintenance in aggregate, because visit frequency rose. Whatever is happening is happening at the point where an owner chooses between a dealer and an independent.

The price account and the evidence against it

The dealership response to defection is ordinarily a price response: discount menus, competitive-price guarantees, coupons against independent rates. The account behind it holds that independents are cheaper, owners are price-sensitive, and share follows price.

The dissatisfaction data does not support it (Cox Automotive 2023). Among owners dissatisfied with the dealership service experience, the stated causes are unexpected costs and poor communication. Unexpected is the operative word, and it describes a property of the process rather than a property of the price.

A price is unexpected when the customer was not told it in advance, or was told a different figure, or was told a figure that changed without explanation. None of those is a statement about the level of the price. A dealer that quotes $900 and charges $900 has not produced an unexpected cost. A dealer that quotes nothing and charges $400 has.

Hien and Long (2024) examine customer responses to service failure through expectation disconfirmation and find that the gap between what was expected and what occurred drives the response, rather than the absolute magnitude of the outcome. Applied here, the finding is inconvenient for the discounting strategy: lowering a price the customer was never told does not close a disconfirmation gap, because the gap is created by the absence of the telling.

The trust series points the same way (Automotive News 2023). Asked why they returned to the selling dealership, 54% of owners gave trust as the reason, down from 62% in 2021. Trust is the word customers use for a supplier whose behaviour they can predict. It degrades when outcomes stop being predictable, which is what an unexpected cost is.

Piha and Avlonitis (2015) study customer defection in retail banking and find defection typically preceded by accumulated dissatisfaction rather than triggered by a single pricing event. The pattern matters here because it implies the dealer usually has warning, and usually does not record it.

What the service department does not know

The information failure has a specific shape, and it is visible in what a typical service department can and cannot answer.

It can answer what was done to a vehicle, because the repair order is a legal and warranty document and survives. It can answer what was charged, because the invoice is an accounting record.

It generally cannot answer what the customer was told before the work began, because the estimate is frequently verbal, given over a phone call, and not written to any record that outlives the conversation. It cannot answer what was declined and why, because declined work is an absence rather than an entry. It cannot answer when this owner is next due, in a form that produces an outbound contact rather than a hope that the owner returns.

Each of those gaps maps directly onto a series above (Cox Automotive 2023). Work that was declined and never followed up is a visit that occurs elsewhere, which is visit volume. An owner never contacted at the due date is a share loss to whoever contacts them first, which is market share. An estimate never recorded is the unexpected cost, which is the stated dissatisfaction.

The gaps are not technology gaps in any interesting sense. A dealer management system records the repair order and the invoice because those are required. It does not require an estimate to exist as a record, a declined line to carry a reason, or a next-due date to generate an action, because none of those is required by anyone outside the business.

What the record would have to hold

The remedy follows from the failure and is narrower than a platform purchase.

The estimate must exist as a record before the work begins, carrying the figure given, the date, and who gave it. That single change converts an unexpected cost into a quoted cost, and it is the one the dissatisfaction data points at most directly.

Declined work must carry a reason and a date. A declined line with a reason is a future service opportunity with a known objection. A declined line without one is nothing at all.

The next service due date must exist as a field that produces an outbound contact, rather than as an inference someone could make from mileage if they thought to. Visit frequency of 2.5 a year (Cox Automotive 2023) means the interval between contacts is roughly five months, which is long enough that an owner will be solicited by someone else inside it.

The service relationship must be attached to the customer rather than to the vehicle. A household that replaces a vehicle has not ended its relationship with the dealer, but a record keyed only on the vehicle identification number behaves as though it has.

None of this requires abandoning the dealer management system, which will continue to be the system of record for the repair order and the invoice. It requires a commercial record alongside it that holds the things the dealer management system was never asked to hold.

What this argument does not establish

No causal claim is made. The evidence here is a correlation between a stated dissatisfaction cause and a measured retention decline, supported by mechanism from the service-failure literature. It is consistent with the information account, and it does not exclude a contribution from price. Independent repair is cheaper for most non-warranty work (Li and Zhang 2024), and some share of the movement is certainly explained by that.

The retention figures are drawn from a single research programme. They are the most widely cited series on dealership service retention, but a second independent source measuring the same quantity would strengthen the argument, and none is cited here because none was found at comparable specificity.

The expectation-disconfirmation and defection literature cited is drawn from airline and retail-banking settings rather than automotive service. The mechanism transfers on the argument that the customer's inference problem is the same, which is an argument rather than a demonstration. A reader who rejects the transfer should treat those two citations as illustrative.

No dealer's results are reported in this paper, and no outcome is attributed to any engagement. The figures are industry series, checkable at the sources given, and the case made from them is a case about what a record should contain rather than a claim about what following the advice has produced.

Conclusion

The defection is real, the segment losing fastest is the one the dealer should hold most easily, and demand for the work has not fallen. Those three facts together locate the problem at the point of choice rather than in the market.

The industry's stated cause is unexpected cost. An unexpected cost is an estimate that was never recorded, never communicated, or never reconciled against the invoice, and each of those is a property of a record rather than of a price. A dealer that discounts an unquoted price has changed the number and left the mechanism intact.

References

Hien, N. N., & Long, N. T. (2024). Customer responses to airline service failure: Perspectives from expectation disconfirmation theory. Sage Open, 14(2). https://doi.org/10.1177/21582440241248334

Li, Y., & Zhang, J. (2024). Omnichannel operations in the automotive aftermarket: Consumer reactions to various strategies. Technology in Society, 79, 102677. https://doi.org/10.1016/j.techsoc.2024.102677

Piha, L. P., & Avlonitis, G. J. (2015). Customer defection in retail banking. Journal of Service Theory and Practice, 25(3), 304–326. https://doi.org/10.1108/jstp-04-2014-0080

Cox Automotive. (2025). Dealerships have lost 12% of service visits to competition since 2018. https://www.coxautoinc.com/insights/new-cox-automotive-study-finds-dealerships-have-lost-12-of-service-visits-to-competition-since-2018/

Cox Automotive. (2023). Service industry study: dealerships losing ground to general repair shops. https://www.coxautoinc.com/insights/2023-service-industry-study/

Automotive News. (2023). Service retention key to dealerships' profitability, study shows. https://www.autonews.com/retail/service-and-parts/an-cox-automotive-service-study-0513/

Conflict of Interest Statement

RevOps HQ is a HubSpot Solutions Partner and is paid to implement CRM and revenue operations systems, including for automotive retail and service businesses. The remedy argued for in this paper is work the firm sells. The argument is constructed from published industry research rather than from the firm's own results, and no client outcome is claimed, but a reader should weigh the recommendation in light of who is making it.

Acknowledgments

Retention, market share and visit frequency series are published by Cox Automotive and reported by Automotive News.