Insights · Measurement

Why Cost Per Lead Is the Wrong Metric for Dealership Advertising, and What to Track Instead

Nick Pascali · Sep 29, 2026 · 7 min read

Short answer: cost per lead fails as a dealership advertising metric for three reasons the data makes plain. Most sales never produce a lead: Cox Automotive and Autotrader traced 875,000 vehicle sales and found 92% had no CRM lead attached. Leads that do arrive rarely close: Urban Science's registration-matched data puts the 30-day close rate on internet leads at 6%, and almost as many buying leads defect to a competitor as buy from the store that received them. And a lead's value depends entirely on its source: phone leads close at more than twice the rate of internet leads, and showroom leads at four times. A metric that treats every lead as equal, counts a tenth of the sales, and stops at the form is not a performance metric. It is an activity metric. The replacements are cost per incremental sale, showroom visits, market share in the primary market area, and sales effectiveness.

The three failures, with the numbers

Most sales are invisible to the CRM. In July 2025 Cox Automotive published a study that matched 875,000 vehicle sales against dealer CRM records. Ninety-two percent of the buyers had never submitted an inquiry that became a lead. The average buyer touched about 62 points of contact during a 95-day shopping window; dealers could see roughly two of them. If a lead report is your scorecard, 92% of your customers do not appear on it. Neither does the media that reached them.

Leads rarely close, and the ones that do often close elsewhere. Urban Science, which matches leads to actual registrations, published 2025 close and defection rates for a 30-day window. Internet leads close at 6% and defect to a competitor at 16%. Phone leads close at 14% and defect at 16%. Showroom leads close at 25% and defect at 20%. In a separate analysis the same firm found that only about one in three leads that buy anywhere buy from the dealership they contacted. So for every internet lead that becomes your sale, nearly three become someone else's.

Lead handling swamps lead cost. Foureyes' 2026 benchmark, drawn from 22,900 dealer websites and 1.4 billion visits, found 42.7% of qualified leads were mishandled and 15.2% were never logged to the CRM at all. About 75% of lead-attributed sales close within three days; a lead that sits over a weekend has lost most of its value. A dealer who cuts cost per lead 20% and mishandles 40% of leads has not improved anything a customer or a manufacturer would notice.

Put the three together and the arithmetic is unforgiving. A $44 search lead (the 2026 WordStream median for automotive sales campaigns) closing at 6% is a $738 cost per sale from that channel, before counting the sales the same media influenced without a form. A $36 social lead closing at 3% is a $1,184 cost per sale. The channel with the cheaper lead produced the more expensive sale. Cost per lead pointed the budget the wrong way.

Why it persists

Cost per lead survives because it is easy. Every platform reports it, every vendor can be judged on it, and it moves in real time. It is also convenient for anyone selling media, because a lead is a result that can be delivered whether or not a car is sold.

The manufacturer does not grade the store on leads. It grades the store on sales effectiveness: actual sales against expected sales in the assigned market area, where expected sales are the brand's regional share applied to the registrations in your territory. The dealer principal does not think in leads either; they think in units, gross, and share. Reporting that stops at the lead form is reporting the marketing department gives itself.

What to track instead

Four numbers, in order of how directly they answer the question "did the advertising sell cars."

1. Cost per incremental sale. An incremental sale is one that would not have happened without the advertising. It is measured by comparison, not attribution: run the media in some ZIP codes or audiences and withhold it from comparable ones, then compare registrations or DMS sales. Google now runs incrementality tests at a $5,000 minimum and Meta added incremental attribution to Ads Manager in 2025, so the tooling is no longer the barrier. The method is in incrementality testing for car dealers. Cost per incremental sale is the only number that can be compared against gross to decide whether a channel earns its budget.

2. Showroom visits. Showroom leads close at 25%, four times the internet-lead rate, which makes a visit the most valuable measurable event before the sale. Location-panel measurement can tie ad exposure to dealership visits for opted-in devices, and a matched control group turns that count into a lift. Cost per measured showroom visit is a useful in-flight metric for reach channels like connected TV, as long as it is paired with a sales outcome rather than used alone.

3. Market share in the primary market area. Registration data from S&P Global Mobility or Urban Science shows every vehicle bought in your territory, including the 92% that never became leads and the ones your competitors sold. Share by ZIP, by segment, and by month is the cleanest scorecard for a growth budget because it cannot be gamed by lead volume and it is the same number the manufacturer uses.

4. Sales effectiveness. The OEM's own metric. Track it monthly, not when the zone manager brings it up. A store at 83 that moves to 95 has done something measurable; a store whose cost per lead fell 20% while sales effectiveness stayed at 83 has not.

Cost per lead does not disappear from the dashboard. It becomes an early-warning gauge for one channel at a time, weighted by that channel's close rate, and never the headline.

What a report built this way looks like

The top line is units and share against the plan. The second line is showroom visits and incremental sales by growth channel, with the holdout comparison shown. The third line is maintenance channels measured on cost per matched sale, using a DMS match-back against search, listing-site, and email conversions. Leads appear on the fourth line, by source, with close rate beside cost so nobody mistakes cheap for good.

That is the format we use in monthly client reviews, and it is the reason we can tell a dealer principal which channel should get the next dollar. The measurement methods behind each line are in how to measure dealership advertising ROI.

Questions dealers ask

Is cost per lead ever a useful dealership metric?

Yes, as a diagnostic inside one channel. If search cost per lead doubles in a month, something changed in the account or the market and it deserves a look. It is not useful for comparing channels or for judging whether advertising sold cars.

What is a good close rate on internet leads for a car dealership?

The 2025 industry figure is 6% within 30 days, from Urban Science's registration-matched data. Foureyes measures 11.7% of sales leads across all sources resulting in a purchase. Both are averages that strong lead handling beats by a wide margin.

How can a dealership measure sales that never became leads?

Two ways. Registration match: compare registrations in exposed ZIP codes or households against unexposed ones using S&P Global Mobility or Urban Science data. DMS match-back: match ad exposure or website visitors against the store's own sales file. Registration data sees the whole market including competitor sales; match-back sees only your own.

Why do manufacturers use sales effectiveness instead of leads?

Because it measures what the franchise exists to do: sell the brand's vehicles in its assigned territory at least as well as the brand sells regionally. It is built on registrations, which capture every sale regardless of how the buyer shopped.

Sources

Next step

Let's put your market on the board.

Tell us about your store and your market. We'll come back with a modeled market share projection and a custom strategy to grow your local share, before a dollar is spent.

Start Your 90-Day Pilot

More insights

Budget