Methodology: how we measure the Dealer Reputation Index
The Dealer Reputation Index ranks the 60 largest car dealer groups in the Netherlands every month on reputation behaviour: what they actually do with their Google reviews. This page explains exactly how we measure that. Every number in the index is counted, not estimated, and this methodology lets you check all of it.
The index is a different measurement from the Automotive Google Review Study. The annual study is the deep analysis: 204,976 full review texts from the 30 largest dealer groups, across their entire history. The index is the monthly thermometer: 60 groups, a rolling twelve-month window, recounted every month. The photo and the film.
The dataset: from 60 groups to 1,882 profiles
The unit of the index is the Google Business Profile. A building with three brand locations counts as three profiles, so the number of profiles per group is a floor for the number of locations, never an overestimate of the group.
For edition 1 (September 2026), the universe consists of 1,882 verified profiles across the 60 largest dealer groups. Every profile was checked in two ways: the website on the profile must point exactly to the group’s official domain (exact domain match, not name matching), and the count per group was compared against the official locations page of all 60 groups. Edge cases were checked by hand. The universe is version controlled (edition 1 runs on v1.0.1), and every change is logged in a changelog.
From those profiles, on the reference date of 28 August 2026, we collected a total of 302,352 public Google reviews, including owner responses and response dates. Of these, 46,861 fall within the measurement window of edition 1.
The window: a rolling twelve months
Each edition measures the twelve full calendar months before the edition month. Edition 1 therefore covers August 2025 up to and including July 2026. The current month never counts: a review from last week may still be in a queue and would wrongly count as unanswered. With every new edition, the window moves forward one month and everything is counted again.
The Reputation Score
The Reputation Score (0-100) combines five ingredients in two groups.
What customers do: the average star rating in the window, and the review volume in the window.
What the group does itself: the response rate (share of reviews with a public reply), the response time (median time between review and reply; reviews with a reply date before the review date are edited reviews and do not count), and the complaint gap (share of 1 and 2 star reviews without a public reply).
We do not publish the exact weighting, so the score cannot be gamed. The ingredients and definitions are fully public. For transparency: these are the same components as the Reputation Score in the Repmanager product. The score measures behaviour you can improve with or without software. We happen to sell the tool for it, which is exactly why everyone is measured identically.
The ground rules
Everyone is measured the same. Groups that are Repmanager customers are measured on exactly the same public Google data as every other group. No labels, no advantage, no removal on request: a single exception would make the whole index worthless.
Dealerships change, so we recount. Groups take each other over, add new brands and close locations. We track mergers and acquisitions in a public register, and before every edition we recount the entire field: every group is checked against its official domain again, so brand additions without a press release are found too.
Honestly incomplete beats quietly wrong. If we know a count is not complete yet, for example because of a recent brand addition or acquisition, that group gets an asterisk (“count being verified”) until the next edition has measured it again. In edition 1, this applies to five groups.
Trucks do not count. The index is about consumer car dealerships. Service points do count: most complaints are about the workshop, and the customer rates the group.
Spotted an error? This is how corrections work
A missing location, a brand that was just added, an acquisition we missed: email [email protected]. We correct demonstrable factual errors visibly in the next edition and note them in the register. We do not adjust scores or remove groups on request.
Frequently asked questions
What is the Reputation Score?
A score from 0 to 100 that combines what customers say (stars, review volume) with what a business does itself (response rate, response time, complaint gap), measured over a rolling twelve-month window on public Google data.
Why don’t you publish the weighting?
Because a public formula can be calculated, and so it can be manipulated. The ingredients and definitions are fully public. The weighting is our work.
How is the index different from the annual study?
The annual study analyses 204,976 full review texts from the 30 largest groups across their entire history. The index counts monthly over a rolling 12-month window for the 60 largest groups. A different measurement with different numbers, both counted and verifiable.
Why does the number of profiles differ from the number of locations?
The index counts Google Business Profiles. One building with several brand locations has several profiles, and not every location has a profile. The profile count is therefore a floor.
