What Should a Dealership Pay Per Lead?

Ask ten dealership marketing agencies what you should pay per lead and you will get ten confident numbers. The confidence is the tell. Nobody who has looked closely at a range of dealership accounts answers that quickly, because a cost per lead is not one number. It is two numbers multiplied together, and most dealerships spend years working on only one of them.

The arithmetic

Cost per lead = (cost per unit of traffic) × (units of traffic per lead).

That is the whole thing. The first term is what you pay to get a shopper onto your site. The second is how many of those shoppers it takes to produce one person who actually contacts the store.

The first term is bought. Media buying moves it — better targeting, better relevance, better auction position. That is our work, and it is where nearly all the industry's attention goes.

The second term is converted. Your website moves it, and nothing else does. No amount of media skill changes how many visitors it takes your site to produce a lead.

Here is why that matters more than it sounds: hold the first term still and your cost per lead moves one-for-one with the second. Same traffic, twice the leads, half the cost per lead. It is not a claim, it is division.

Most dealers only work on one term

Almost every conversation a dealer has about lead cost is about the first term. Cheaper clicks, better keywords, a new channel, a lower CPM. Those are real levers and we pull them — Wheeler clients see cost per click two to six times lower than category benchmarks, which is the single clearest thing we can show about the media side.

But if you have negotiated your traffic costs down for three years and your website still converts the way it did in 2019, you have been optimising one half of a multiplication and leaving the other half alone.

The second term is usually the larger opportunity, for a simple reason: nobody measures it. Dealers know their cost per click. Very few know how many vehicle page views it takes their site to produce one lead.

What the second term actually looks like

Our partner platform, Scout, published measurements on exactly this. Across 3.37 million vehicle page views and 16,142 leads, dealer websites on the older platforms took an average of 210 vehicle page views to produce one lead. The same measurement on Scout came to 108.

The range on the legacy side is the part worth sitting with: from 161 views per lead at the best end to over 900 at the worst. Scout publishes the low end of that range rather than the flattering one.

Two hundred and ten views to one lead, against a hundred and eight. Same shoppers, same markets, same media rates. Roughly double the leads out of traffic the dealer had already paid for — and therefore roughly half the cost per lead, without renegotiating a single thing on the media side.

That is the term nobody is working on.

Finding your own number

The reason there is no benchmark worth quoting is that "lead" does not mean the same thing at any two dealerships. Before you compare yourself to anyone, answer three questions about your own store.

1. What counts as a lead here?

Write it down before you open any report. A reasonable definition: a person who contacted the dealership about buying a unit. Not a service appointment, not a parts enquiry, not a four-second phone click, not the same person filling out two forms. If your definition and your agency's definition differ, every performance conversation you have is two people talking past each other.

2. Does your analytics actually count that?

Open your ad account and read your list of conversion actions. Most dealers never have. You will typically find a stack of them, several set up by vendors who are long gone, at least one duplicating another. Anything on that list that is not your definition from question one is inflating your lead count and flattering your cost per lead.

3. Does it reconcile with your CRM?

This is the check that settles it. Take last month. Count the leads your CRM actually received. Compare that to what the ad platform claimed. If the platform says four hundred and the CRM has ninety, the platform is not measuring leads, and nothing else you do with the reporting matters until those two numbers are in the same neighbourhood.

When dealers run that reconciliation honestly, the real cost per lead is usually a good deal higher than the reported one. That is uncomfortable, and it is also the most valuable thing you can learn about your marketing, because it is the first number in the chain that is true.

Why we report vehicle page views

There is a reason we report on Vehicle Detail Page views rather than leaning on a conversion column.

A VDP view is someone clicking into a specific unit to look at photos, pricing and specs. That is a shopper doing what a shopper does. It is hard to fake and hard to accidentally configure, and it moves in step with sales rather than in step with your tracking setup. A website session, by contrast, can be somebody who hit the homepage and left.

We would rather report a number that only goes up when something real happened, including in the months when a friendlier number was available.

The short version

There is no benchmark cost per lead worth quoting. There is an equation with two terms, and you almost certainly have someone working on the first one and nobody working on the second. Define what a lead means at your store, audit what your account actually counts, reconcile it against the CRM — and then look hard at how much traffic your website is spending to produce each one.

If you would rather someone else ran that audit, ours is free.