Marginal CAC: Price the Next Customer, Not the Average One

Short answer: Marginal CAC is the cost of the next customer you get from the next dollar of spend. Blended CAC averages every customer together, so it can look healthy even when the last chunk of budget bought very few customers. Scale while marginal CAC stays below the contribution margin of a first order (or your target payback), and hold or cut when the next customer costs more than they are worth.

These ideas come from RetentionX's newsletter; we've summarized them and added how we use them in our Meta work. The brands in the examples are made up.

Free download: Marginal CAC, the one page summary (PDF). Print it or share it with your team.

What is marginal CAC?

Marginal CAC is the change in spend divided by the change in net new customers between two spend levels.

CAC (customer acquisition cost) is what you pay to get one new customer. Most brands track it as an average: total spend divided by total new customers. That is blended CAC. The RetentionX framing is simple: blended CAC tells you whether the machine is healthy, and marginal CAC tells you whether to feed it more.

Why is average CAC misleading when you scale?

Because the average hides how expensive the last customers were.

The newsletter's illustrative example: a brand spends $50,000 and gets 1,000 new customers. Blended CAC is $50. Looks fine. But the last $5,000 of that spend may have bought only 60 customers. That last slice cost about $83 per customer.

Now say, for illustration, a new customer is worth $70 in first order margin. Then the average says "keep going" and the margin says "you are already losing money on the last slice." The average blends cheap early customers with expensive late ones, and the expensive ones are exactly the ones you are deciding whether to buy more of.

This is how diminishing returns work on any ad platform. The first dollars go to the easiest buyers. Each extra dollar has to work harder to find someone new.

How do you calculate marginal CAC?

Compare two spend levels and divide the difference in spend by the difference in new customers.

Marginal CAC = change in spend / change in net new customers

Two details matter:

  1. Count new customers from your store's orders, not Meta attributed conversions. Meta's numbers include customers who would have bought anyway and can be inflated by remarketing.
  2. Count net new customers only, meaning first time buyers, not repeat orders.

Here is a made up example. Nightfold sells a $38 weighted sleep mask. All numbers are illustrative. Last week it spent $10,000 and got 250 new customers in Shopify, a blended CAC of $40. This week it raised spend to $13,000 and got 300 new customers. Blended CAC is about $43, which looks like a small change. But the extra $3,000 bought only 50 extra customers. Marginal CAC is $60. That is the number that should decide the next raise.

What is the decision rule?

Scale while marginal CAC is below the contribution margin per first order, or below your target payback. Hold or cut when the next dollar costs more than the customer is worth.

Contribution margin is what a sale leaves you after product cost, shipping, fees and discounts. If your business only works when the first order pays back the ad cost, compare marginal CAC to first order contribution margin. If you can wait for repeat purchases, compare it to the margin you collect within your target payback window.

Notice what the rule is not: it is not a target ROAS. RetentionX makes this point directly, and we agree. ROAS (return on ad spend) is an average world metric. It will approve spend that the margin math rejects, because a healthy average can sit on top of a loss making last slice.

How do we use marginal CAC in our Meta workflow?

This is our own application of the idea. We use it in four places.

The daily scaling gate

Before any budget raise, we check what the last raise bought. If the previous increase moved blended CPA (cost per acquisition) by more than the size of the raise itself, the curve is bending, and the next raise waits. This fits into how we promote winners to Scaling: a winner earns more budget, but the account as a whole has to show that the extra money is still buying customers at a price that works.

The weekly check

Each week we compare spend and net new customers against the prior week and calculate marginal CAC on the difference. We pair it with net new reach, meaning how many people saw the ads for the first time. Flat reach on rising spend is a leading indicator: the account is paying more to reach the same people, and marginal CAC tends to rise after it.

Client reports

We put blended and marginal side by side. In plain words: the average customer costs $X, the next one costs $Y, and here is where we scale or hold. It gives clients a clear reason why we sometimes recommend not spending more, and it explains why CAC rose when the budget doubled. That second question comes up a lot, and marginal CAC is the honest answer.

Big sale periods like Black Friday

During BFCM we set budget tiers in advance, each with criteria for pulling back. Then we check marginal CAC in short windows of a few hours, compared against the same time blocks on earlier sale days. Our shorthand: the 2pm raise has to earn the 6pm one. Sale days move fast, and a raise that is not buying customers wastes money quickly.

How can you make marginal CAC sharper?

RetentionX lists several ways to go further once the basic version is in place:

  • Correct it for incrementality. Incrementality means sales your ads actually caused, not sales they took credit for. Lift tests or geo holdouts let you turn attributed marginal CAC into incremental marginal CAC. Our guide to the Meta conversion lift test covers how those tests work.
  • Build a spend response curve. With 12 or more weeks of spend and new customer data, fit a curve that shows diminishing returns. The slope of that curve at your current spend is your marginal CAC.
  • Break it down by layer. Calculate it by campaign, audience or concept to see which parts of the account still have room.
  • Compare it to marginal LTV. Buyers at the top of the curve often skew toward promos and repeat less. The real ceiling is marginal CAC against marginal LTV (lifetime value of those specific customers), not against your average customer.
  • Use prospecting only. Remarketing inflates the customer count because it reaches people who were already likely to buy. We explain why in is your remarketing incremental.

What mistakes do teams make with marginal CAC?

  • Judging scale on blended CAC or ROAS alone. Both are averages and both hide the last slice.
  • Using Meta attributed conversions. Count first time buyers from your store.
  • Comparing periods that are not alike. A sale week against a normal week will distort the math.
  • Reading one day as a trend. Daily numbers are noisy. Use the daily check to pause a raise, and the weekly view to make bigger calls.
  • Raising budget again before the last raise has been measured. Each raise should earn the next.

Key takeaways

  • Marginal CAC is the cost of the next customer at the next dollar of spend.
  • Blended CAC tells you if the machine is healthy; marginal CAC tells you whether to feed it more.
  • Calculate it as change in spend divided by change in net new customers from store orders.
  • Scale while marginal CAC is below first order contribution margin or target payback; hold or cut above it.
  • Do not use target ROAS as the scaling rule; it is an average and hides the last slice.
  • Pair it with net new reach: flat reach on rising spend usually comes before rising marginal CAC.

FAQ

What is the difference between blended CAC and marginal CAC? Blended CAC is total spend divided by total new customers, an average of every customer you bought. Marginal CAC is the cost of only the extra customers you got from extra spend. Blended tells you about health; marginal tells you about the next decision.

Why did my CAC go up when I doubled my budget? Because the extra money had to find harder to reach buyers. The first dollars buy the easiest customers. Each extra dollar costs more per customer, which pulls the average up.

How much data do I need? You can start with two weeks of spend and new customer counts. For a proper spend response curve, the newsletter suggests 12 or more weeks.

Should marginal CAC include remarketing? Ideally no. Remarketing reaches people who were already likely to buy, so it makes new customer counts look better than they are. Use prospecting spend and first time buyers.

Sources: RetentionX newsletter, "Marginal CAC: pricing the next customer, not the average one" (concept, illustrative example, decision rule and ways to sharpen it).

Find out what your next customer really costs

If you want to know whether your account has room to scale or is already paying too much for the last slice, we offer a free audit. You get a full account audit plus net new reach and cumulative reach reports. Request it on our contact page.

Check what you learned

Tap an answer to see if you got it right.

1. What is marginal CAC?

Blended CAC hides the last slice.

2. When should you keep scaling?

Hold or cut above it.

3. How do you calculate it?

Use store orders for customers.

Ready for the full test? Take the Retention exam and get certified. Want this run on your account? Get my free audit.

About the author

Jake Mehani, founder of Creative Analytics

Jake has spent 15 years running Meta and paid social ads for consumer brands. Meta published his RBX Active Advantage+ case study, he taught the DMAnc Meta and TikTok ads masterclass, and he spoke at Motion's 2024 Creative Strategy Summit. His senior team runs media buying, creative and measurement like an in house team.

Jake on LinkedIn  |  Case studies  |  Free Meta audit

Previous
Previous

How Much of Your CAC Is Just Maintenance? Growth Spend vs Replacement Spend

Next
Next

Is Your Growth Real? Cohort Quality and Customer Equity Dilution