CAC Is the Wrong Hill: The 2 Metrics That Predict Channel Health

In the last few videos I gave you the AI stack, and then the memory agent that runs on top of it. Those are tools. This is the scoreboard you're judging them by.

Customer acquisition cost is the number almost every operator I know leads with. It's on the dashboard, it's in the weekly standup, and it's the first thing anyone quotes when they defend a channel.

CAC can fall every month while the business underneath it gets weaker, and your dashboard will congratulate you the whole way down.

Why do I say that? Because CAC only measures one side of the transaction, and the side it measures is the side you can manipulate.

Key Takeaways

• CAC is the easiest number in your business to move, which is exactly why it's the easiest one to fool yourself with. Cheaper clicks, harder discounts, and demand you were going to capture anyway all push it down.
• LTV to CAC has a floor and a ceiling. Below 3 to 1 the channel doesn't feed itself. Above 7 to 1 you're under-investing in something that's already working.
• Payback period is the metric that decides whether you can fund next month. Under 12 months for consumer, 18 to 24 for B2B on multi-year contracts.
• The 2 metrics answer different questions. The ratio asks whether the economics work. Payback asks whether your cash can wait for them. A channel can pass one and fail the other.
• Calculating LTV takes about 5 minutes. Average order value, times average orders per customer, times gross margin. Use margin, not revenue.
• Run both numbers on your 2 biggest channels before you plan the quarter. One of them will surprise you.

The easiest number to move

CAC is what you pay to land one new customer. Total acquisition spend divided by new customers acquired, over whatever window you're looking at.

You can drop it tomorrow. Chase the cheapest clicks, discount harder, or go after people who were already going to buy from you anyway.

Your CAC goes down. Your business gets weaker. I've watched teams defend a channel for 3 quarters because that number kept ticking down.

That last one is worth sitting with. Retargeting your own warm list, bidding on your own brand terms, running a discount to people mid-decision: all of it produces a beautiful CAC and almost none of it produces incremental customers. You're paying to take credit for demand you already had.

What CAC leaves out

2 things, and they're the 2 that matter.

The first is worth. A $400 CAC is a bargain against a customer who returns $3,000 in margin and a disaster against one who returns $500. The cost figure alone can't tell you which you bought.

The second is time. A channel that returns its acquisition cost in 8 months and a channel that takes 34 months can post identical CAC and identical lifetime value.

One of them funds your next quarter. The other one is a loan you're making to yourself.

So the fix is 2 metrics, run together, on every channel you spend real money on.

Metric 1: LTV to CAC

Lifetime value is what a customer is worth to you over their whole relationship with you. Divide that by what you paid to get them, and you have a ratio.

3 to 1 is the working floor. Below 3 to 1, the channel doesn't feed itself: the margin one customer returns won't cover acquiring the next one plus the overhead sitting around them.

Above 7 to 1, you're being too careful and leaving growth on the table. A channel returning 7x its acquisition cost can absorb more spend than you're giving it, and holding it at that efficiency usually means you're capping volume to protect a number.

The band between those 2 numbers is where a healthy channel lives. Treat it as an operating range with a floor and a ceiling.

How to calculate LTV in 5 minutes

Average order value, times average orders per customer, times your gross margin.

That's it! No data team, no attribution model, no warehouse query.

Use gross margin rather than revenue. Revenue you never keep can't pay back an acquisition cost, and running the ratio on revenue is how a channel that looks like 6 to 1 turns out to be 2 to 1 once cost of delivery comes out.

And if you don't know your average orders per customer, pull your last 20 invoices and count. Close enough beats not knowing at all.

Metric 2: payback period

Payback period is how many months before a customer pays back what you spent to get them.

This is the one that decides whether you can fund next month. The ratio tells you if a channel works eventually. Payback tells you if you're still standing when eventually shows up.

Under 12 months if you sell to consumers. 18 to 24 months if you sell to businesses on multi-year contracts. Longer than that, and the channel is borrowing against a future you have to survive to reach.

The arithmetic is simpler than the ratio. Take the CAC, divide it by the gross margin that customer generates per month, and you have your answer in months.

The payback math, worked

Here's where the 2 metrics stop agreeing with each other, which is the entire reason you run both.

Say you're looking at 2 channels. Same spend, same CAC, and a cost-per-lead dashboard that shows them dead even.

Channel A: paid search.

• CAC: $600
• Average order value: $450
• Average orders per customer: 4
• Gross margin: 60%
• LTV: 450 x 4 x 0.60 = $1,080
• LTV to CAC: 1,080 / 600 = 1.8 to 1
• Monthly margin per customer: roughly $54 (4 orders spread across 20 months)
• Payback: 600 / 54 = about 11 months

Channel B: partner referral.

• CAC: $600
• Average order value: $1,200
• Average orders per customer: 3
• Gross margin: 65%
• LTV: 1,200 x 3 x 0.65 = $2,340
• LTV to CAC: 2,340 / 600 = 3.9 to 1
• Monthly margin per customer: roughly $65 (3 orders spread across 36 months)
• Payback: 600 / 65 = about 9 months

Read those side by side. Identical CAC, identical spend, and a 2.2x gap in what each customer is actually worth.

Channel A fails the floor. At 1.8 to 1 it's returning less than 2 dollars of margin for every dollar of acquisition, which doesn't cover the next customer plus the overhead. It pays back inside a year, so it doesn't feel like a problem, and that's precisely what makes it dangerous: healthy cash timing on unhealthy unit economics reads as a working channel right up until you scale it.

Channel B clears the floor at 3.9 to 1 and pays back in 9 months. That's the one to put money behind, and the cost-per-lead view would never have told you.

Now change one input on Channel B. Stretch those 3 orders across 72 months instead of 36, which is what happens when a referral partner sends customers who buy slowly.

Monthly margin drops to about $33, and payback stretches to 18 months. The ratio is untouched at 3.9 to 1. The channel just became a financing decision.

That's the case the ratio alone hides, and it's why the 2 numbers run together.

Where this goes wrong

4 failure modes I run into most often.

Running the ratio on revenue instead of margin. The single most common error, and it inflates every channel by whatever your cost of delivery is. A 60% margin business running LTV on revenue overstates the ratio by 67%.

Treating LTV as a forecast. It's an estimate built on averages, and averages hide mix. If 20% of your customers drive 80% of the repeat volume, blended LTV describes a customer who doesn't exist. Segment it before you bet a quarter on it.

Assuming payback is linear. Most businesses collect unevenly, with a bigger first order and a long tail. If your first purchase covers half the CAC, your real payback is much shorter than the monthly-average math implies. Use actual collection timing when you have it.

Ignoring the blended-versus-incremental gap. Brand search, retargeting, and email to existing contacts all post spectacular CAC because they harvest demand other channels created. Judge them on incremental customers, not attributed ones, or your best-looking channel is just the best-attributed one. Our incrementality guide walks through how to separate them.

What to do before you plan the quarter

Pick your 2 biggest channels by spend. Run both numbers on each one. That's the whole assignment, and it's 30 minutes of work.

One of them will surprise you. I've seen 2 channels sit 3x apart on the back end while the cost-per-lead dashboard showed them dead even.

Then change what the dashboard leads with. CAC is a real number and it earns its spot. Put these 2 above it.

That's the difference between a channel that looks cheap and a channel that pays you back.

Further Reading

On Professor Leads

• CPR Calculator runs your spend, leads, and revenue through cost per lead, cost per revenue dollar, and ROI in one pass. It's the fastest way to see how far apart your cost view and your revenue view really are.
• Incrementality Guide separates the customers a channel created from the ones it only took credit for, which is the gap that makes a low CAC lie.
• Your Q2 Reality Check is the quarterly review this post feeds into, and where these 2 metrics belong in it.
• Your Lead Gen Strategy Is Broken is the strategic layer sitting underneath the measurement layer.

On Forbes (by William DeCourcy)

• Why Chasing Metrics Is Killing Your ROI (And How To Fix It) makes this same argument at the portfolio level: what happens when a team optimizes the metric instead of the outcome.
• The Leadership Metric You're Not Tracking (But Should Be) is on picking the measure that predicts rather than the one that's easy to collect.