Your LTV Is Fiction: The Fully-Burdened Math That Should Gate Your Ad Spend

The myth goes like this. Your customers are worth $400 over their lifetime, the gospel ratio is 3x, so you can happily pay up to ~$130 to acquire one and still be sitting pretty.

Here's the reality. That $400 is revenue, over a timeframe nobody has actually defined, before you've subtracted the cost of the product, the shipping, the refunds, the discounts, the payment fees, or the chargebacks. The number you can genuinely afford to spend isn't ~$130. It's often closer to ~$45. And the gap between those two figures is exactly why brands that look "profitable" in the dashboard quietly run out of cash.

I've watched this play out enough times to believe it's the single most common way good ecommerce brands talk themselves into unprofitable scaling. Not weak creative. Not bad targeting. A lifetime value number built on hope.

So let me give you the worksheet we use to rebuild a client's real LTV before we'll agree on a single CAC target. It's five steps. None of them are clever. They're just the steps most people skip.

1. Pick a time horizon, and make it boring

The first thing that makes LTV fiction is that nobody bounds it. "Lifetime" quietly means "forever", and forever is where you can justify any CAC you like.

So the first move is to put a fence around it. I'd use three years. Not the heat death of the universe - three years. That's a window you can actually model, a window your cash flow actually lives in, and a window where the cohort data isn't pure guesswork.

Why does this matter so much? Because the further out you push the horizon, the more lifetime value you get to "count", and the more you let yourself overpay today against revenue that may never turn up. The DTC brands that got into trouble a few years back nearly all made the same mistake - they banked on future LTV that never materialised, went long on payback, and discovered too late that they'd been buying customers at a loss the whole time.

A three-year LTV you can defend beats a ten-year LTV you invented. Pick the boring number.

2. Use delivered-to-door gross profit, not revenue

This is the big one, and it's where most worksheets fall apart.

When you say a customer is worth $400, ask the follow-up: $400 of what? Almost always the answer is revenue. But revenue is the number before your business has paid for anything. It's not what you keep.

What you actually want is delivered-to-door gross profit. As in: by the time the product is physically sitting on the customer's doorstep, how much money did you genuinely take out of that order? That means starting from revenue and stripping out the full cost of getting the thing there.

Here's roughly what that looks like on a single order with a ~$70 AOV:

• Revenue: ~$70
• Cost of goods: ~$21
• Shipping and fulfilment: ~$11
• Payment processing (around 3%): ~$2
• Discount actually used (the code they redeemed): ~$7

That ~$70 order is now closer to ~$29 of gross profit before you've spent a cent acquiring them. So the customer who "generates $400 of revenue" over three years might only be handing you ~$165 of actual delivered-to-door profit. Same customer. Less than half the number.

If you build your CAC ceiling off the $400 and not the ~$165, you will overspend on every single customer and the P&L will tell you the truth about six months later than you'd like.

3. Subtract the leaks nobody puts in the model

Step two gets you delivered-to-door profit per order. But there's a second layer of money that walks out the door across the relationship, and it almost never makes it into the spreadsheet. This is what "fully burdened" actually means - you scrape every part of the business and ask where the little leaks are.

The usual suspects:

• Refunds and returns. If 8% of orders come back, that's not just lost revenue, it's the shipping you ate both ways plus the cost of any product you can't resell. On a returns-heavy category this can quietly remove a tenth of your gross profit.
• Chargebacks and fraud. Small as a percentage, real in dollars, and almost always missing from the LTV calc.
• Customer support and the cost to serve. If a slice of customers email three times before they're happy, that's a real cost attached to those orders.
• Gift cards and store credit that get issued and never fully redeemed in the way you modelled.

None of these are dramatic on their own. Together they routinely shave 10-20% off the gross profit you thought you had. So that ~$165 customer is now maybe a ~$140 customer.

To put this into perspective: the difference between modelling a customer at ~$165 and at ~$140 doesn't sound like much. But run it across 20,000 customers a year and it's the difference between a comfortable margin and a board meeting about the burn rate.

4. Decide if you're even an LTV business

Here's a hard one that a lot of founders don't want to sit with. LTV only earns its place in your CAC maths if customers genuinely come back.

I've seen plenty of brands quote a healthy three-year LTV when, in reality, their repeat purchase rate is under 15%. If that's you, your "lifetime" value is basically your first-order value with a rounding error attached. And in that case the honest metric isn't LTV-to-CAC at all. It's first-order profit to CAC. You need to make a real margin on the very first purchase, because there probably isn't a second one coming.

So before you let LTV carry your acquisition budget, look at one number: what percentage of customers buy again within twelve months? If it's strong - say north of 35%, or you've got a genuine subscription - then LTV is real and you can lean on it. If it's thin, model on the first order and treat any repeat as a bonus, not a plan.

My take: it's far safer to build your CAC ceiling so the first order roughly pays for itself, and let repeat revenue be the upside. Brands that survive the bad months are the ones that didn't need the second purchase to break even.

5. Now apply the 3x bar, and watch your "profitable" target move

Now you've got a real number. Three-year, delivered-to-door, post-refunds, post-everything. Let's say it lands at ~$140.

The best-in-class rule of thumb is a 3x LTV-to-CAC ratio. So against ~$140, your CAC ceiling is ~$47. That's the most you can pay to acquire a customer and still be running the disciplined economics that keep a business fundable and alive.

Now compare that to where you started. The myth said the customer was worth $400 and you could spend ~$130. The real worksheet says the customer is worth ~$140 and you can spend ~$47.

That's not a rounding difference. That's a ~$130 target versus a ~$47 target - and if you've been buying customers at ~$90 thinking you had headroom, you've actually been losing roughly $43 on every one. Volume was making the hole bigger, not smaller. This is the moment the worksheet usually goes quiet in the room, because a target everyone assumed was conservative turns out to be roughly triple what the unit economics can carry.

The reason the 3x bar matters isn't that the number is magic. It's the buffer. That cushion is what absorbs the customers who churn early, the cohorts that underperform, the month Meta costs jump. Spend right up to a 1.5x and the first bad cohort puts you underwater.

What this does to how you scale

Once you've got a real CAC ceiling, scaling actually gets simpler, not harder.

The discipline becomes: I will grow as fast as the platform lets me at that ceiling, and not a dollar faster. If the market hands me 4,000 profitable customers this month, I take them. If it's only 1,500, I take 1,500 and I don't loosen the ceiling to hit a vanity revenue number. The ceiling is the thing everyone in the business should be able to recite - founder, head of growth, the person approving the ad budget. When it's vague, it drifts upward every time growth stalls, and that drift is where the losses live.

And it reframes what "profitable" even means. A brand can be profitable per the dashboard's ROAS column and still be unprofitable per its real economics, because the ROAS column is counting revenue the business never actually kept. Build the ceiling off burdened LTV and the two finally agree.

So here's what I'd do this week. Take your most-quoted LTV figure - the one in your pitch deck or your head - and run it through these five steps. Bound the horizon. Strip it to delivered-to-door gross profit. Subtract the leaks. Sanity-check your repeat rate. Then divide by three.

If the CAC ceiling that falls out is lower than what you've been paying, you've just found the leak before it found you. If you'd rather not rebuild the model alone, a Signal/Noise Audit walks through exactly this - your real unit economics, your true LTV, and the honest CAC your account can actually carry - so you're scaling on numbers that survive contact with the P&L. Either way: when you redo the math, where does your real ceiling land?