In offices in a lot of cities there are water coolers, and every week or two someone arrives with fresh bottles.
If you’ve ever worked in such an office, try to picture that person. Most people can’t. They arrive, swap the bottles, get a signature and leave. They don’t chat. They don’t linger. They don’t get drawn into whatever is happening in the room. It’s a small masterpiece of not being noticed.
That reads like ordinary professionalism. It isn’t. It’s arithmetic.
The economics of water delivery are brutally tight, and the whole business rests on one fact: acquiring an office costs more than six months of delivering to it. Half a year of bottles, drivers, fuel and invoices just to get back what was spent winning the account.
Which means everything after the sale is designed around not losing it. Don’t be memorable. Don’t be a topic. Don’t be present at an awkward moment. Do the job and leave, and in three years nobody will ever have considered switching.
And if a contract does end at month three, the company didn’t make a small loss on that account. It would have been better off never delivering at all — every bottle it carried was carried at a loss, in service of a payback that never arrived.
That’s this whole idea in one image: the most dangerous number in your business is the one you’ve already spent and haven’t earned back yet.
The number nobody calculates
The three maximums — per customer, per lead, per click — answer how much.
There’s a second question, and almost nobody asks it: when?
acquisition cost ÷ margin per customer per month = payback period
A cleaning company acquiring a recurring client for $50, earning $51 a month from that client, has a payback of about one month. That business can advertise almost as fast as it can find customers.
A business acquiring a customer for $300 against $50 a month gets six months. Same margin per customer, same lifetime value — completely different business from a cash perspective. One can accelerate at will. The other is lending money to its own growth, at scale, in advance.
Now the number that actually matters:
monthly acquisition spend × payback period ≈ the cash you need before you start
Spending $4,000 a month on customer acquisition with a six-month payback means that at full tilt you’ll have something on the order of $24,000 sitting out in the world, unrecovered, at any given moment.
That figure is deliberately conservative — the true number is somewhat lower, because early cohorts start paying back while later ones are still out. Use the conservative version anyway. Nobody has ever gone under from being too careful about this particular number.
If you don’t have that money, you don’t have a marketing plan. You have a countdown.
What lifetime value actually is
Now the multiplier.
How long a customer stays, how often they buy, and what else they bring you — multiply them out and you often find a customer is worth three, five, seven times their first purchase.
One thing to be exact about, because the error is expensive: the multiplier applies to profit, not to revenue. A customer who spends $265 four times has not brought you $1,060 to spend on acquisition. At a 50% margin they’ve brought you about $530, and that’s the number your maximum is built from. Anybody quoting a lifetime value in revenue is quoting a figure they cannot spend.
Count the relationship across everything you sell, too, not the single product — or every contractor comes out worth exactly one job.
The temptation is obvious. Multiply your maximum by five and go spend it. Your competitors are bidding against first-sale margins; you’d be bidding against three years of profit. You’d win every auction in your market.
You’d also, quite often, destroy the business.
Lifetime value is self-referential
That multiplier of five isn’t a fact about your customer. It’s a fact about your customer and your business together, and it holds only if you keep being the business that earned it.
Trace what happens when you spend it in advance.
The money goes out this month. The returns arrive over three years. Ninety days in, you’re short. Being short doesn’t stay financial for long: you can’t cover payroll comfortably, so you lose two crew members. Down two people, you’re late on jobs. Late jobs get done in a hurry. Rushed work produces complaints, and complaints produce reviews.
Now look at what you assumed. You assumed the customer would still be with you in year three. You assumed they’d buy repeatedly. You assumed they’d recommend you. Every one of those assumptions was really an assumption about quality — and quality was the first thing to go when the cash ran short.
You spent a lifetime value and, in spending it, made it impossible.
That’s what makes this different from ordinary financial risk. An aggressive bet that doesn’t pay off is bad luck. This one isn’t a bet at all — it’s a loop. The spending removes the conditions that justified it.
Which is why the multiplier isn’t a number you apply. It’s a number you qualify for.
The question
So before anyone multiplies anything, one question, asked out loud:
Do you have the cash to survive the wait?
Not “will it be worth it.” Not “do the numbers work.” Those are the two-point test and the answer hasn’t changed. This is a different question, and it’s about your bank account between now and the payback date.
If yes — use the multiplier
You have working capital that covers monthly acquisition spend multiplied by the payback period, with room left over. Then use it, because you now have a structural advantage: you can outbid competitors who only count their first sale, and it isn’t recklessness on your part, it’s a better-informed bid.
Use the figure you actually observe. If your records say a customer buys 1.5 times, use 1.5 — don’t shade it down to feel safer. A number you’ve invented a margin of safety into is no longer a measurement, and you’ll have lost track of which part of it is real. The safety isn’t in shrinking the multiplier. It’s in the cash test you just passed.
One condition attached, though. Retention is now a financial control, not a service metric. If repeat rates fall, your maximum falls with them, and you’ll be spending against a lifetime value that no longer exists. Review it quarterly, not annually — and if the number moves, recompute before your next budget decision rather than after it.
One more thing, and it lands hardest in the trades where people leave easily — cleaning, appliance repair, handyman work. There you’re not only betting that the customer stays three years. You’re betting that the person who serves that customer stays too. If the cash gets tight and the crew earns less than it expected, it goes — and it takes the repeat business with it. How easily your people can leave is an input to this decision.
If no — price off the first sale
Advertise only what pays for itself now. Compute your maximums from average job, margin and conversion rates, and stop there. Don’t multiply.
This is not the timid version. It reads like the cautious option and it isn’t — it’s the correct calculation for a business without a cash cushion. A business that prices off first sale grows more slowly and survives. A business that prices off lifetime value without the capital to bridge it grows faster right up until it doesn’t.
And there’s a hidden benefit: a channel that pays for itself in month one is a channel you can turn off in month two with no consequences. That optionality is worth a lot when you’re small.
Two ways through when the answer is no
Find channels that need less capital. Not every source of customers requires paying up front for each one. Referral work, existing-customer marketing, search optimization, local listings — slower, but they don’t ask you to finance every acquisition six months ahead. When cash is the binding constraint, the cheapest customer isn’t the one with the lowest cost per lead. It’s the one who doesn’t need to be prepaid.
Start small enough that the gap stays controllable. You don’t have to choose between full speed and nothing. Run at a volume where the unrecovered amount stays inside what you can absorb, and let it grow as returns start landing. Slower, and it never puts the business at risk.
Both are downgrades from what you’d do with capital. Both are enormously better than the alternative, which is discovering your payback period in the same week you discover you can’t make payroll.
The connection back to quality
One more thing, because it changes how the whole idea reads.
The water delivery driver who says nothing and leaves quickly isn’t being polite. He is protecting a payback period. His invisibility is a financial instrument.
Look at your own business and you’ll find the same thing in reverse: whatever protects your repeat business is protecting your maximum. A repairman who does a $200 job properly can earn close to a decade of work from it. Not because he’s principled — or rather, he is, but it also happens to be the highest-return financial decision available to him that day, and he probably never calculated it.
So when the cash gets tight and something has to give, notice what you’re actually choosing between. Cutting quality to protect cash flow reduces the lifetime value that justified your acquisition costs, which lowers your maximum, which means you can afford fewer customers, which tightens cash further.
Same loop, entered from the other side.
Which is why the answer to a cash squeeze is almost never “spend less on delivering the work.” It’s “buy fewer customers for a while.” One of those decisions is reversible next quarter. The other takes years to repair, if it repairs at all.
What to write down
- Your cost to acquire one customer
- The margin that customer produces per month
- Payback period: the first divided by the second
- Monthly acquisition spend × payback period — roughly the cash this plan requires, deliberately on the high side
- Whether you have it
Five lines. If line five is no, you have your answer, and it’s the second outcome of the two-point test wearing different clothes: not no, but not at this speed.
This chapter is drawn from Win in the Spreadsheet First. The seven inputs behind the maximums are in the calculator, and the full method is in the book.
Frequently asked questions
- What is a payback period and how do I calculate mine?
- Take what you spend to acquire one customer. Take the margin that customer produces per month. Divide the first by the second, and you have how many months your money is gone before it comes back. A cleaning company acquiring a recurring client for $50 and earning $51 a month from them has a payback of about one month — that business can advertise almost as fast as it can find customers. A business acquiring a customer for $300 against $50 a month gets six months. Same lifetime value, completely different business from a cash perspective: one can accelerate at will, the other is lending money to its own growth, in advance, at scale.
- How much cash do I need before I start advertising?
- Multiply your monthly acquisition spend by your payback period. Spending $4,000 a month with a six-month payback means that at full tilt you'll have something on the order of $24,000 sitting out in the world, unrecovered, at any given moment. That figure is deliberately conservative — the true number is somewhat lower, because early cohorts start paying back while later ones are still out. Use the conservative version anyway. Nobody has ever gone under from being too careful about this particular number. If you don't have it, you don't have a marketing plan. You have a countdown.
- Can I bid against lifetime value instead of the first sale?
- Only if you can survive the wait. Lifetime value is real — a customer is often worth three, five, seven times their first purchase — but that multiplier isn't a fact about your customer. It's a fact about your customer and your business together, and it holds only if you keep being the business that earned it. Spend it in advance and trace what happens: the money goes out this month, returns arrive over three years, ninety days in you're short, you lose two crew members, jobs run late, rushed work produces complaints, complaints produce reviews. Every assumption behind the multiplier was really an assumption about quality, and quality was the first thing to go. You spent a lifetime value and, in spending it, made it impossible.
- Should I apply lifetime value to revenue or to profit?
- Profit, and the error here is expensive. A customer who spends $265 four times has not brought you $1,060 to spend on acquisition. At a 50% margin they've brought you about $530, and that's the number your maximum is built from. Anybody quoting a lifetime value in revenue is quoting a figure they cannot spend. Also count the relationship across everything you sell rather than the repeat rate on a single product — a remodeler who counts per-product lifetime value concludes every customer is worth exactly one job and prices himself out of his own market.
- What do I do if I don't have the cash cushion?
- Price off the first sale: compute your maximums from average job, margin and conversion rates, and stop there. Don't multiply. This isn't the timid option, it's the correct calculation for a business without a cushion — one that prices off first sale grows more slowly and survives, one that prices off lifetime value without the capital to bridge it grows faster right up until it doesn't. Two ways through: find channels that don't need capital per customer (referrals, existing-customer marketing, search optimization, local listings), and start small enough that the unrecovered amount stays inside what you can absorb.