Calculating true customer acquisition cost across dealer and digital channels

Three ledgers paid to sell the car on that driveway, and none of them can tell you what it cost. The manufacturer ran national brand and offer advertising. The dealer bought local search, direct mail and a marketplace subscription. The marketplace spent to rank on Google for "used SUV near me", then charged the dealer for the introduction it produced. One retailed unit, three sets of books, no shared denominator.

Customer acquisition costCustomer acquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition →, or CAC, is the total marketing and sales spend needed to produce one retailed unit. Writing the formula takes ten seconds. Deciding whose spend and whose unit belong in it is the actual job, and it is where most automotive CAC numbers fall apart.

What counts inside CAC (and what does not)

Include:

• Paid digital media (search, social, listing sites)
• Lead and listing purchases from marketplaces such as Autotrader or Cars.com
• Traditional and co-op advertising: radio, direct mail, local TV
• The marketing-attributable share of sales staff time and showroom cost
• Agency fees and martechmartechThe connected set of software tools a marketing team uses to plan, run, measure and automate campaigns across channels.View full definition → (CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition →, chat, call tracking, attributionattributionA framework for assigning credit to the touchpoints that contributed to a conversion, so you can measure which channels and interactions actually drive results.View full definition →)

Exclude:

• The vehicle itself, which is cost of goods
• F&I (finance and insurance) product costs
• Overhead with no role in attracting buyers (accounting, service bay tooling)

Co-op advertising is shared spend where the OEM (original equipment manufacturer) reimburses the dealer for part of a campaign. Only the dealer's net out-of-pocket share belongs in that dealer's CAC. A $10,000 campaign at 50% co-op is $5,000 of dealer CAC, and the other $5,000 belongs to the OEM's number rather than to nobody's.

The contested line is customer cash. Industry trackers have put average US incentive spend per new vehicle in the low thousands of dollars in recent years, which dwarfs most media budgets. Rebates are normally treated as price, so they sit in revenue and gross, not in CAC. That is defensible. What is not defensible is excluding incentives and then celebrating a falling CAC that was bought with $2,000 on the hood. Pick a treatment, write it down, keep it for at least four quarters.

The core formula

$$\text{CAC} = \frac{\text{Total acquisition spend for a channel}}{\text{Retailed units attributed to that channel}}$$

Retailed units, not deliveries, not leads. Fleet and wholesale volume never belongs in the denominator, because marketing did not buy it. Neither do units that unwind: Carvana sells with a seven-day return window, and any retailer running a return policy has to net returned vehicles out of the denominator or it will report a CAC it never achieved.

Step 1: split spend by channel

Build a clean monthly spend table. Direct costs are easy; shared cost is where judgment enters.

Floor traffic is not free. The salaried greeter, the lit and heated sales floor, the desk manager working a deal: a share of that is acquisition cost for walk-in buyers. Allocating the marketing-attributable portion of sales-floor labour and facility cost to the walk-in channel holds up under audit, because turning bodies into orders is what a showroom is for.

Step 2: attribute units to channels

A buyer who clicked a Google ad and then walked in is not a walk-in. Two simple models:

• Last-touch credits the final interaction. Easy, and it overcredits the showroom.
• First-touch credits whatever created the lead. Better for judging what fills the top.

Most dealers settle on lead-source attribution: whichever channel produced an identifiable lead (form fill, tracked call, marketplace handoff) gets the unit, and walk-ins with no prior tracked touch count as showroom. That only works if identities stitch across web session, phone and CRM record. Customer data platformsCustomer data platformsSoftware that unifies customer data from every source into one persistent profile that marketing, sales and service teams can act on.View full definition → sell exactly this capability (Segment, part of Twilio, is one), and the licence fee belongs inside CAC, which the finance team will point out before you do.

Google's explainer on attribution models in Analytics is a clean primer on the logic.

A worked example

One mid-size US franchise dealer, one month. Illustrative figures for teaching, not benchmark data.

Digital channel:

• Paid search and social: $18,000
• Third-party leads: $9,000
• Martech share: $3,000
• Total digital spend: $30,000
• Retailed units attributed to digital: 50

$$\text{Digital CAC} = \frac{30{,}000}{50} = \$600 \text{ per unit}$$

Showroom / walk-in channel:

• Co-op radio and direct mail, net of reimbursement: $12,000
• Signage and local brand: $4,000
• Marketing-attributable sales-floor labour and facility share: $26,000
• Total showroom spend: $42,000
• Retailed units attributed to walk-ins: 30

$$\text{Showroom CAC} = \frac{42{,}000}{30} = \$1{,}400 \text{ per unit}$$

Digital looks cheaper. Two cautions before anyone moves budget.

Blended CAC hides the marginal number

The $600 averages every dollar already spent. The next $10,000 into paid search will not buy 16 more units, because the cheap high-intent queries are already bought and you are climbing the bid curve into broader ones. The number that governs a spend decision is marginal CAC: incremental spend divided by the incremental units it produced. Hold everything else still, run the increase for a full shopping cycle, measure the delta. Dealers who scale on the blended figure find out the marginal one was $1,200 after the quarter has closed.

Second caution: $600 and $1,400 mean nothing on their own. Read them against the multi-vehicle value figure the LTVLTVLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.View full definition → lesson builds, and against front-end gross, which at US franchise dealers has typically run in the low thousands of dollars per unit. A $1,400 walk-in CAC leaves that front end close to hollow, so the deal only works if service, parts and the next replacement cycle show up.

Whose CAC is it: OEM, dealer, marketplace

Add the three ledgers together and you get system CAC per retailed unit, a number nobody reports. It runs well above any single participant's figure, for two structural reasons.

Non-exclusive leads. A marketplace can sell the same shopper's enquiry to several dealers in one market. Each pays, one delivers the car. The losing dealers' spend was real and produced nothing, so cost per retailed unit at market level sits far above the winning dealer's. As a dealer your only defences are close rate and speed of first response, which the funnelfunnelThe customer journey from awareness to purchase, typically Awareness, Interest, Consideration, Decision, Action, with prospects narrowing at each stage.View full definition → lesson instruments. As an OEM, your defence is making your own channels the cheaper route to the same shopper.

Cost shifting. When an OEM trims national tier-1 spend, dealer CAC rises: less demand arrives pre-sold, dealers buy more marketplace leads, competition for a finite pool of shoppers lifts lead prices, and system cost per unit climbs while the manufacturer's marketing line improves. The OEM dashboard shows a win the network is paying for. Any OEM CMO tracking only factory spend per unit is flying on one instrument.

Carvana is the useful counter-example. One company holds the media spend, the lead, the close and the delivery, so advertising expense divided by retail units sold is an auditable CAC anyone can compute from its reporting, and it has moved by hundreds of dollars a unit between its growth years and its cost-discipline years. Franchise networks cannot produce the equivalent without agreeing shared definitions across parties who each have reasons to prefer their own.

Agency sales models in Europe rearrange the same accounting: when the manufacturer owns the transaction and pays the retailer a handling fee, most acquisition spend lands on the OEM P&L and dealer CAC falls for reasons unrelated to efficiency. Normalise for that before comparing your figure to any external range, using the credibility tests the benchmarks lesson sets out.

Common mistakes that corrupt CAC

Counting leads, not retailed units. Cost per lead flatters digital. Divide by units that actually retailed.

Ignoring co-op reimbursement. Gross campaign cost overstates dealer CAC. Use net out-of-pocket.

Treating the showroom as free. Zeroing sales-floor and facility cost makes digital look expensive and hides the real trade-off.

Matching this month's spend to this month's units. Shopping cycles run from a couple of weeks to a couple of months. A month where you doubled spend will look brilliant or disastrous purely on timing. Cohort by lead date and wait for the cohort to close.

Double-counting cross-channel buyers. One attribution model, applied consistently. Inconsistency is worse than an imperfect model.

Blending new and used. The margins and lead sources differ sharply. Segment them.

Putting it into a decision

• If digital CAC is $600 but those buyers rarely come back to the service drive, the answer is post-sale work on that cohort, not more leads at the same price.
• If showroom CAC is $1,400 and those buyers finance and service, that channel earns more co-op-funded local advertising, not less.
• If the marketplace bill keeps rising while your close rate is flat, you are subsidising a non-exclusive lead. Renegotiate on exclusivity or move the money to channels you own.
• Recompute monthly and by campaign, because lead prices and close rates swing with inventory and incentives.

The target is not the lowest CAC. It is the widest defensible spread between what a unit costs to acquire and what the customer behind it is worth.

Key Takeaways

• CAC = acquisition spend divided by retailed units from that channel. Net out returns, fleet and wholesale, and use co-op spend net of OEM reimbursement.
• The showroom is not free. Allocate the marketing-attributable share of sales-floor labour and facility cost to walk-ins, or you will wrongly condemn digital.
• Blended CAC is a report; marginal CAC is a decision. Test an incremental spend increase over a full shopping cycle before scaling anything.
• Three parties pay for one unit. Sum OEM, dealer and marketplace spend to see system CAC, and watch for cost shifting that improves one dashboard while raising the network's true cost.
• Decide the incentive question once. Whether customer cash sits in CAC or in gross, keep the treatment stable for four quarters or your trend line is fiction.