Dealership Marketing Attribution: From Clicks to Cars Sold

Here's the uncomfortable number I open with when a dealer principal asks me whether their marketing works: in most stores, less than 10% of sold units can be tied to any marketing source in the CRM. One widely cited analysis put it at 8%. The other 90-plus percent of your sales show up as "walk-in," "repeat," or a blank field — while your agency's report claims credit for all of them.

Both things can't be true. After thirty years in operations, I can tell you the gap isn't because marketing doesn't work. It's because attribution at most dealerships was never built — it was assumed. This is the system I install for automotive clients to close that gap, and you can build most of it with tools you already own.

Why the CRM misses 90% of the story

A modern car buyer racks up hundreds of digital touchpoints over 60–90 days — searches, third-party listings, your VDPs, streaming ads, social — and then walks in. Your CRM captures exactly one moment: the lead form or the showroom up. Everything before it evaporates. Three structural problems cause this:

1. Last-touch logic. The CRM credits whatever created the record. If a customer saw your conquest campaign for six weeks and then Googled your store name, Google search gets the credit and the campaign that did the work gets nothing. Branded search is almost always over-credited for this reason.

2. The walk-in black hole. Forty-plus percent of buyers in many stores never become a digital lead before showing up. No click, no form — so digital reporting can't see them, even though digital media put them in the car.

3. The agency grades its own homework. Each vendor reports its own platform metrics — impressions, clicks, "engagements" — with no connection to your DMS. Add up the sales every vendor claims and you'll find you sold 400% of your actual volume.

The fix: matchback against the DMS

The technique that cuts through all of it is sales matchback: instead of asking "what did this click do?", you start from the sold log and work backwards.

The mechanics:

1. Export the sold file from your DMS monthly — buyer name, address, email, phone, VIN, deal date. Service ROs too, if you're measuring fixed ops.
2. Hash and match. Customer identifiers are hashed (privacy-safe, household-level) and matched against the exposure files from each marketing channel — the device IDs your geofencing vendor captured, the emails your CRM mailed, the households your CTV campaign reached.
3. Apply an attribution window. A sale counts toward a campaign only if exposure happened within a defined window — 30–90 days is the standard for vehicle sales, shorter (30 days) for service. Define it in writing, because vendors will quietly stretch it to claim more.
4. Report sold units by channel, with overlap shown honestly. A buyer touched by three channels should appear as a multi-touch sale, not three separate sales.

Done right, this gets you VIN-level attribution: this campaign, this buyer, this unit, this gross. That sentence ends more bad agency relationships than any audit ever will.

Choosing an attribution model without overcomplicating it

You'll hear about first-touch, last-touch, linear, time-decay, U-shaped. Here's my operator's shortcut: for a dealership, the model matters less than the matchback. Whether you weight touches linearly or by recency changes channel credit by a few points; failing to match sales at all changes it by an order of magnitude.

My practical recommendation:

• Run last-touch in the CRM (you already do — keep it for lead-handling accountability).
• Run matchback monthly for true sold-unit attribution by channel.
• Layer multi-touch only when your monthly matchback shows heavy channel overlap and you genuinely need to arbitrate budget between two big channels. About half of marketers have moved to multi-touch; for a single-point store, monthly matchback plus common sense gets you 90% of the value at 10% of the cost.

And for big-budget channels, add the occasional incrementality check — hold out a comparable market area or audience for 60 days and compare sold rates. Some buyers were coming anyway; a holdout is the only way to know how many.

What this looks like channel by channel

• Paid search: conversion tracking is table stakes; matchback exposes how much "conversion" volume is branded search harvesting demand other channels created. Break branded and non-branded into separate campaigns or you'll never see it.
• Geofencing and programmatic display: demand the device-level exposure file and a conversion-zone visit report, then matchback to the DMS. I detail the full setup in my geofencing for car dealerships breakdown — including why $15–$25 per verified visit is the benchmark.
• CTV/OTT: household-level exposure makes CTV one of the easier channels to matchback, despite being "unclickable." Ask for household match rates, not completion rates.
• Third-party listings: matchback their lead file, but also matchback their shopper exposure claims. The delta is usually educational.
• Email/CRM marketing: simplest of all — you own both files. If your agency can't tie email sends to sold units, that's a willingness problem, not a technical one.

The payoff, in numbers

Stores that run integrated attribution aren't just better informed — they're more profitable. Industry analyses peg the difference at roughly 27% less wasted ad spend and materially higher per-unit profitability, simply because budget migrates from channels that report well to channels that sell cars. With the average dealer now spending about $739 per vehicle sold on advertising, a 27% efficiency gain on a 100-unit store is real money — every month.

The first step costs nothing: pull last month's sold file, pull every vendor report, and count how many sold customers any vendor can actually identify. If the answer embarrasses everyone, start with my dealership ad spend audit guide — it's the checklist version of this article — or have my team at ADT Consulting's automotive practice run it with you.

FAQ

What is sales matchback in automotive marketing?

Matchback is the process of matching your DMS sold file (names, emails, addresses, VINs — privacy-safe hashed) against marketing exposure files to determine which sold units were touched by which campaigns within a defined attribution window. It works backwards from sales instead of forwards from clicks.

What attribution window should a dealership use?

30–90 days between marketing exposure and the sale is standard for vehicle purchases, reflecting the typical buying journey; 30 days is typical for service. Define the window contractually with every vendor — an undefined window lets everyone claim every sale.

Why does my CRM show so few sales from marketing?

CRMs capture only the final conversion point — the lead or the up — and miss the dozens of prior touchpoints, which is why studies find under 10% of sales tied to marketing sources in standard CRM data. Matchback against the DMS recovers the missing attribution.

Is multi-touch attribution worth it for a single dealership?

Usually not as a first step. Monthly DMS matchback with a defined window delivers most of the decision-making value. Add multi-touch modeling when matchback shows significant overlap between major channels and you need to split budget between them.