ROAS for Dealerships: The Only Metric That Actually Matters

1. The Metric Hierarchy

Every dealership marketing conversation involves metrics. Impressions. Clicks. Leads. CPL. CTR. The problem isn't a lack of metrics — it's that everyone's focused on the wrong ones.

Here's the full metric hierarchy for dealership advertising, ordered from least meaningful to most meaningful:

Most dealerships — and virtually all agency reports — stop at Level 3. They know how many leads came in and what each lead cost. Everything below that line is either unmeasured, manually tracked in a spreadsheet, or estimated based on gut feeling.

But the metric that actually determines whether your advertising budget is making or losing money sits at Level 7. Gross profit. Not revenue — gross profit. That's the number that pays the bills, funds the next month's campaigns, and determines whether your ad spend was an investment or an expense.

The metric hierarchy isn't complicated. Everyone agrees that sold units matter more than clicks. The problem is that most dealerships have precise data on the metrics that don't matter and zero data on the metrics that do.

Here's the uncomfortable reality: a dealership can have a declining CPL, rising lead volume, impressive click-through rates, and still be losing money on advertising. How? Because the leads aren't turning into deals. Or they're turning into low-gross deals on vehicles you shouldn't be advertising. Or they're turning into deals that would have happened anyway through organic traffic. The metrics at the top of the hierarchy can look great while the metrics at the bottom are terrible.

Conversely, a campaign with a "terrible" CPL of $150 might be producing buyers who close at 25% with $4,000 in average front-end gross. That campaign is printing money — but the CPL-focused report would flag it for budget cuts.

The metric hierarchy exists. Every GM intuitively understands it. The problem is that the tools, vendors, and reporting systems in automotive retail are all optimized for the top of the hierarchy — and almost nothing measures the bottom.

2. What ROAS Means in Automotive

ROAS — Return on Ad Spend — is a simple formula with a critical nuance for automotive retail.

The standard formula:

ROAS = Revenue Generated ÷ Ad Spend

That works for e-commerce. A store spends $1,000 on ads, generates $5,000 in sales, and the ROAS is 5x. Clean and useful.

But automotive isn't e-commerce. A dealership doesn't keep the full sale price. A $45,000 deal might have $3,500 in front-end gross — or it might have $800. The revenue number is almost meaningless without the gross profit number.

Automotive ROAS Formula

Automotive ROAS = Total Gross Profit from Campaign Deals ÷ Campaign Ad Spend

This is the number that matters. Not revenue. Not topline. Gross profit — the money you actually get to keep after cost of goods.

Why Revenue-Based ROAS Is Misleading

Consider two campaigns, each with $5,000 in ad spend:

Illustrative example showing why revenue-based ROAS is misleading in automotive.

Campaign A looks dominant on revenue — 36x ROAS. Four deals. $180,000 in sales. But those were mini deals. Low gross. High pressure from price-focused buyers. After cost of goods, the campaign barely broke even at 1.04x gross profit ROAS.

Campaign B generated fewer deals but attracted buyers willing to pay for value. Three deals at an average of $4,200 gross each. Gross profit ROAS of 2.52x — more than double Campaign A's actual return.

Revenue-based ROAS would tell you to scale Campaign A. Gross profit ROAS tells you to scale Campaign B. The difference between those two decisions — made every month for a year — is the difference between a marketing budget that compounds and one that bleeds.

What to Include in "Gross Profit"

For campaign-level ROAS, most GMs should start with front-end gross — the profit on the vehicle itself, before F&I products. Here's why:

• Front-end gross is directly influenced by the quality of the lead and the campaign targeting. A campaign that attracts payment buyers will produce different front-end gross than one that attracts cash buyers.
• F&I backend is valuable but influenced more by the F&I manager's skill than by the campaign that generated the lead. Including it can obscure campaign-level signals.
• Service absorption and long-term customer value are real but too delayed and too noisy to attribute at the campaign level in real time.

Start with front-end gross. As your data matures, layer in F&I. But never use revenue — it tells you almost nothing about whether the campaign made money.

3. Why CPL Is a Vanity Metric

CPL — Cost Per Lead — is the single most overused metric in dealership advertising. It's the headline number on every agency report. It's the metric GMs ask about first. And it's almost completely useless as a measure of advertising effectiveness.

That's a strong claim. Here's why it's true.

CPL Treats All Leads as Equal

A "lead" in automotive can mean a lot of things:

• A serious buyer who's been researching for three weeks, has a trade-in, and wants to schedule a test drive this Saturday
• A price shopper who submitted forms to six dealerships and will buy from whoever quotes the lowest number
• Someone who clicked accidentally on mobile and auto-filled a form
• A competitor's employee checking your inventory and offers
• A bot or scraper that filled out the form with fake data

CPL treats all five of those as identical. A $30 CPL that's producing Category 2-5 leads is worse than a $120 CPL that's producing Category 1 leads. But the CPL report says the $30 campaign is winning by a mile.

CPL Optimizes for the Wrong Thing

When your agency optimizes for lowest CPL — which is what most agencies do, because it's what most GMs ask for — they're optimizing for the cheapest attention. That means broader targeting (more reach, lower quality), more aggressive creative (clickbait headlines, unrealistic offers), and more permissive lead forms (fewer fields, lower barrier).

All of those tactics lower CPL. All of them also lower lead quality. You get more leads, each one costs less, and fewer of them buy anything. The agency report looks great. Your showroom traffic doesn't change.

The Real Cost of a "Cheap" Lead

Consider the hidden costs of low-quality leads:

• BDC or AI time: Every lead gets a response — a text, an email, a phone call. Low-quality leads consume the same resources as high-quality ones.
• Salesperson time: If a low-quality lead does book an appointment (rare), the salesperson spends 30-60 minutes with someone who was never going to buy.
• Opportunity cost: While your team is chasing cheap leads, they're not following up on the expensive-but-serious leads that would actually close.
• Morale: Nothing burns out a BDC or sales team faster than weeks of leads that never respond, never show, and never buy.

A $30 CPL that produces 100 leads and 1 deal costs $3,000 per deal. A $120 CPL that produces 25 leads and 5 deals costs $600 per deal. The "expensive" leads are five times cheaper when you measure what matters.

CPL measures how cheaply you can buy a name and a phone number. ROAS measures how effectively you can turn ad dollars into gross profit. They're not the same thing — and optimizing for one often destroys the other.

4. How to Calculate True Automotive ROAS

Here's a step-by-step walkthrough with illustrative numbers. This isn't a real dealership — it's designed to show exactly how the calculation works and where the data comes from.

Step 1: Isolate the Campaign

Pick one campaign. Let's say it's a Silverado Spring Event running on Meta for 30 days.

Step 2: Count the Leads

How many leads did this specific campaign generate? Not "how many leads from Facebook" — how many from this campaign, with this creative, this targeting, this budget.

That CPL looks high. Hold that thought.

Step 3: Track Engagement

Of those 35 leads, how many actually engaged with your follow-up? This is where lead quality starts to reveal itself.

A 74% response rate on a $120 CPL campaign. Compare that to a $30 CPL campaign where 20% respond. The quality difference is already visible.

Step 4: Count Appointments and Shows

Step 5: Count Deals and Gross

Step 6: Calculate ROAS

All figures above are illustrative, designed to demonstrate the ROAS calculation methodology.

For every $1 spent on this campaign, the dealership generated $5.43 in front-end gross profit. The cost to acquire each sold unit was $700. That $120 CPL that looked expensive? It produced deals at $700 per sale. A $30 CPL campaign that produces 1 deal out of 100 leads costs $3,000 per sale — four times more expensive when you measure what matters.

The Data You Need

Notice what's required to run this calculation:

• Campaign-level lead attribution: Not "Facebook" — the specific campaign that generated each lead
• Conversation data: Whether the lead responded and engaged
• Appointment data: Whether an appointment was booked and whether the customer showed
• Deal data: Whether the deal closed and the gross profit amount

That's four data points from four different systems. If any one of them is missing or disconnected, you can't calculate true ROAS. You're stuck with CPL.

5. The Benchmarks

Once you can calculate ROAS, you need context. What's good? What's average? What means you should kill the campaign immediately?

General ROAS Benchmarks for Automotive

Benchmarks are illustrative ranges based on industry observation. Individual dealership results vary based on market, brand, and competitive dynamics.

These ranges assume you're measuring against front-end gross. If you include F&I backend, ROAS numbers will be higher across the board — but the relative ranking of campaigns stays the same.

Channel-Level Benchmarks

Different advertising channels tend to produce different ROAS profiles. Here's what the data typically shows:

Typical ranges are illustrative, based on observed patterns across dealership campaigns. Actual results depend on market conditions, creative quality, and lead handling effectiveness.

Notice the pattern: the channels with the lowest CPL (Meta prospecting) tend to have the lowest ROAS. The channels with higher CPL (Google Search non-branded, Meta retargeting) tend to have higher ROAS. This is exactly why CPL is a poor metric — it often points you in the opposite direction of profitability.

The Lead Handling Multiplier

One critical factor the benchmarks above don't capture: lead handling quality massively affects ROAS. The same campaign producing the same leads will generate wildly different ROAS at two different dealerships — one with sub-60-second AI response and persistent follow-up, the other with a 4-hour response time and one unanswered call.

A dealership with excellent lead handling (immediate AI response, 5-step follow-up cadence, appointment confirmation, show reminders) can push a 3x ROAS campaign to 6x simply by converting more of the leads that are already coming in. The campaign didn't change. The creative didn't change. The budget didn't change. The handling changed.

That's why ROAS visibility without lead handling optimization is only half the picture. You need to see both — which campaigns produce good leads and how effectively your team or AI converts those leads into deals.

6. How to Get ROAS Visibility

By now, the case for ROAS over CPL should be clear. The question is: how do you actually get ROAS visibility when your data lives in five disconnected systems?

Option 1: Manual Stitching

You could try to do this manually. Pull the campaign report from your agency. Export leads from the CRM. Cross-reference against appointment logs. Check the DMS for closed deals. Map each deal back to a campaign.

In theory, this works. In practice, it doesn't — for three reasons:

1. The data doesn't match: Your agency says 47 leads from "Facebook." Your CRM shows 52 leads from "Internet." Which 47 out of the 52? Which campaign? Nobody knows, because the campaign-level identifier doesn't survive the handoff between systems.
2. It's always late: By the time someone manually compiles this data — usually 2-3 weeks after month-end — the budget decisions for next month have already been made based on the agency's CPL report.
3. Nobody maintains it: Manual processes work for a month or two. Then the person responsible gets busy, the spreadsheet gets stale, and everyone reverts to CPL.

Option 2: Third-Party Attribution Tools

There are third-party attribution platforms designed to stitch data across systems. They integrate with ad platforms, CRMs, and sometimes DMS systems. They're better than manual tracking — but they have limitations:

• They don't own the conversation: If your AI tool or BDC system isn't integrated, there's still a gap between "lead generated" and "appointment booked."
• CRM source tags are unreliable: Attribution tools rely on the CRM's source field, which is often inaccurate or too generic to provide campaign-level insights.
• DMS integration is hard: Most DMS systems are notoriously difficult to pull data from. Integration is expensive, fragile, and often delayed.
• It's another vendor: You're adding a sixth system to stitch together the other five. Each integration point is another potential data gap.

Option 3: Closed-Loop Architecture

The structural solution is a system where demand generation and lead handling run through the same data infrastructure. When the company that creates the campaign also handles the AI response, the appointment booking, and the outcome tracking — the attribution chain is continuous. No handoffs. No stitching. No data gaps.

Here's what that requires:

• Campaign ID on every lead: Not "Facebook" — the specific campaign, ad set, and creative that generated the lead. UTM parameters carried through from click to close.
• Conversation data linked to campaigns: The AI response system knows which campaign produced each lead and uses that context in the conversation.
• Appointment data linked to campaigns: Every booked appointment carries the campaign identifier from the original lead.
• Deal value linked to campaigns: When a deal closes, the gross profit data connects back to the originating campaign.

That's closed-loop attribution. It's the only way to get true ROAS visibility at the campaign level — and it's the only way to stop making $50,000+/month decisions based on a metric (CPL) that doesn't tell you whether you're making money.

ROAS is the metric. Closed-loop attribution is how you get it. Everything else — CPL, CTR, impressions — is just noise that sounds like signal.