Car Dealer Lead Generation: Every Channel by Cost per Sale
Our founder also runs a car-dealer marketing agency: $1,500 retainer plus $1,500 ad spend, and a typical month produces three to five sales. That's $600 to $1,000 per sale. The same inventory posted to Facebook Marketplace through autobook.io Pro, at $99 a month, prices out between $9 and $24 per sale.
Those are our numbers, not a benchmark we found somewhere. The agency belongs to autobook.io's owner. Publishing the cost per sale of your own paid-media business is an odd thing to do, and it's the only version of this article worth writing, because the alternative is running the comparison against numbers we'd have to invent.
How do you price a sale, exactly?
One line of arithmetic:
Everything the channel costs in a month ÷ the units you can honestly attribute to it.
The division is easy. The honesty is the hard part. Most stores track advertising as one lump — a total, a percentage of gross, a line the GM defends in the monthly meeting — and never break it down to a per-channel cost per unit. The NADA Data report publishes dealership financial benchmarks every year, including what franchised stores spend on advertising, and it's worth comparing your own store's number against it. But an industry average won't tell you which of your channels is carrying the others.
Two rules before you run the numbers. Count the whole cost, including what you pay someone to manage the spend — the retainer is part of the channel, not overhead. And count only the units you'd actually defend in front of the person who wrote the check.
What does a $3,000-a-month agency actually produce?
Here is the real structure, from that agency, running paid social for car dealers:
| Month | Retainer | Ad spend | Total | Sales | Cost per sale |
|---|---|---|---|---|---|
| Typical | $1,500 | $1,500 | $3,000 | 3–5 | $600–$1,000 |
| Best ever | $1,500 | $1,500 | $3,000 | 10–11 | $273–$300 |
That is not a bad agency result. Somebody builds the creative, sets the targeting, watches the spend, and reports on it. The buyers it produces are people who weren't shopping ten seconds before the ad interrupted them, which is genuinely hard to do and genuinely valuable. Three to five deals a month for $3,000 keeps a small store's paid program alive.
It is also, per sale, expensive. And on the agency's best month in its history — eleven sales, everything firing — it still lands at $273 a sale.
Where does Marketplace posting price out?
autobook.io Pro is $99 a month and includes 300 credits. Credits are the unit of work:
- Import a vehicle: 2 credits
- Post it to Marketplace: 1 credit
- AI-written description: 1 credit
- AI photo re-render: 2 credits per image — premium and opt-in, applied to the vehicles you choose
Across the platform, roughly 11% of posted vehicles sell — about one in nine. Run those two facts together and the cost per sale depends entirely on how much work you buy per vehicle:
| What you run per vehicle | Credits | Vehicles on 300 credits | Units sold at 11% | Cost per sale |
|---|---|---|---|---|
| Import + post | 3 | 100 | ~11 | ~$9 |
| Import + AI description + post | 4 | 75 | ~8 | ~$12 |
| All of the above + 2 AI images | 8 | ~37 | ~4 | ~$24 |
Nine dollars to twenty-four. Against $600 to $1,000, that's roughly 25 to 100 times cheaper per sale. Against the best month the agency has ever had, it's still somewhere around 11 to 30 times cheaper.
Esad at Napleton Buick GMC put the same math in salesperson terms: "one sale, I make my money for the month." He also described the subscription as "quarter of a car sale." Carlos at Bruceland Buick GMC said the cost was "so low it'd be absurd" not to use it, and reported "18 cars sold month to date, 5 of those from autobook.io, plus two more buyers it brought in off the street."
So what's the catch?
Three of them, and none is small.
The 11% is a platform-wide average, not your number
It's every posting dealer on autobook.io divided into every sale reported back. That pool includes people who post daily and answer every message inside ten minutes, and people who imported once in March and forgot about it. Your ratio will be your own, and it will move with your inventory, your market, and how fast you reply. Treat 11% as the shape of the thing, not a forecast. Pro is sized for eight to twelve extra sales a month, which is plan sizing, not a promise.
The subscription isn't the real cost — your time is
A rep working Marketplace properly spends about 40 minutes a day answering buyers. Call it 13 to 17 hours a month. There is no inbox inside autobook.io. No CRM, no AI replies, no message templates in the product, no notifications. Every one of those conversations happens in your own Facebook Messenger, typed by you.
That's the honest ledger: the software removes the grind — importing, writing, photo work, posting, renewing — and leaves the job. Put your own hourly number on 15 hours and add it to the $99 before you compare anything. Agency buyers need working too, so nobody's eating free here. But the $9 figure buys you posting, not selling.
They aren't the same buyers
Paid social interrupts someone scrolling. Marketplace catches someone already typing "Silverado" into a search box. Those are different people at different points, and a store that turns off its paid program on the strength of this table is making a decision the table doesn't support. The claim here is cost-efficiency, not replacement. We sell the Marketplace tool, so read that sentence with the appropriate suspicion — and then go run the numbers on your own invoices, which is the only version that matters.
How do you rank your own lead generation channels?
We're not publishing a portal price table. Portal rate cards aren't public, they vary by market, package tier and whatever your rep negotiated, and a made-up number in a table like this would poison everything above it. What we'll give you instead is the worksheet.
- Pull three months of invoices, one line per channel. Portals, paid social, agency retainers, third-party listing fees, the tool subscriptions. Whole cost, including management.
- Ask every buyer where they found the vehicle and write it on the deal sheet. Crude, and usually more accurate than the attribution report, because the buyer is the only one who knows.
- Divide, per channel, per month. Three months smooths out one lucky week.
- Rank them and look at the bottom.
| Channel | Monthly cost | Units attributed | Cost per sale |
|---|---|---|---|
| Portal / classifieds subscription | |||
| Paid social (spend + management) | |||
| Purchased buyer inquiries (third-party vendors) | |||
| Marketplace posting | $99 | ||
| Walk-ins, referrals, repeat | $0 |
Two things tend to fall out of that exercise. The channel with the biggest invoice is rarely the one with the best cost per sale. And the free channels get no credit at all, because nothing arrives to remind you they exist — there's no invoice for a Marketplace listing, so nobody defends it in the meeting. If the worksheet says Marketplace should be running across the whole floor rather than one desk, the staffing and per-rep cost side is in the dealership owner's playbook.
On the attribution side, autobook.io does one specific thing: when you mark a listing sold inside Facebook, that signal reports back and the platform works out whether the sale is attributable to autobook.io. That's detection and credit. It is not a reporting dashboard — there's no per-listing revenue screen to log into. The aggregate network numbers are published live on our comparison pages, and they update automatically. No other vendor in this category publishes a post-to-sold ratio at all, which is worth knowing if you go looking for one to compare against.
What does running it look like day to day?
You paste your dealership website URL and the importer pulls the inventory. Most dealership website platforms are supported, plus CarGurus, Cars.com and AutoTrader, and a generic crawler for anything else. A 100-vehicle store usually takes five to ten minutes, sometimes longer. A typical import brings in the VIN, price, mileage, trim, colors and a median of 20 photos per vehicle.
Posting runs through a Chrome extension in your own browser, from your own logged-in account, on your own IP. Maximum 50 vehicles per batch. It paces itself, types at human cadence, and skips a vehicle and retries when Facebook throws an error. Renewing existing listings is in the extension too — it scans your listings page, shows you a review screen, and clicks renew where Facebook offers it.
Daily volume should follow your account's history, not the batch ceiling. Never listed on Marketplace before: two to five a day. Posted before, even manually: 10 to 15 from the start. Established account that's run a posting tool: 15 to 30. That's operating experience rather than a published Facebook rule — the longer version is in how many cars you can post on Marketplace.
Photo work stays optional on purpose. It replaces backgrounds with studio scenes, removes plates and plate frames, adds reflections and shadows, and preserves the actual vehicle including aftermarket modifications. It classifies interior shots and leaves them alone, because buyers are meticulous about the space they'll sit in. At 2 credits an image it's the line item that moves your cost per sale most, so spend it on the units that deserve it.
Capacity by plan: 100 active listings on Pro, 235 on Elite, unlimited on Ultimate. If you run out of credits mid-month, top-ups are 100 for $20, 250 for $45, 500 for $80. autobook.io is US-only today. And the standard caveat, which is not boilerplate: you're responsible for the accuracy of your listings and for following Facebook's Commerce Policies and your local dealer-advertising rules. The listings post from your account. The compliance is yours.
Questions people actually ask
Is $9 per sale a guarantee?
No. It's arithmetic on a platform-wide ratio and a published price. Your inventory, your market and your reply speed will move it. The honest version of the claim is that the ceiling here — $24 a sale with full AI descriptions and photo work — sits far below the floor of a managed paid-media program.
Should I cancel my agency and just post to Marketplace?
Not on the strength of this article. Paid social reaches people who weren't looking; Marketplace reaches people who were. If your paid program is producing sales at $600 a piece and your gross per unit is comfortably above that, it's working. The argument is to add the cheap channel, then decide with three months of your own numbers.
How many credits does one vehicle actually consume?
Import and post is 3 — 2 to import, 1 to post. Add 1 for an AI description. Each AI-generated image is 2 more. So a fully worked vehicle with two re-rendered photos is 8 credits, and Pro's 300 credits cover about 37 of those.
What if my dealership's website platform isn't supported?
Tell us and we'll add it, usually within 24 hours. Support is email, with a 24-hour response standard.
Does autobook.io reply to buyers for me?
No. There's no inbox, no AI replies and no templates in the product. You work Facebook Messenger yourself. That's the 40 minutes a day, and it's the part that stays yours.
What does it cost to start?
Plans from $99 a month — Pro at $99, Elite at $199, Ultimate at $299, with yearly at $990, $1,990 and $3,588. Every number in this article runs off Pro and its 300 credits.
If you want the channel explained rather than priced, start with the complete Marketplace guide for dealers, or the salesperson's playbook if the account in question is yours rather than the store's.