When you choose a business model, you are also choosing when money arrives relative to when it leaves. That timing — not your margin, not your TAM, not your growth rate — decides what will kill you. Two companies can post identical income statements, the same revenue, the same margin, the same growth, and one compounds into a fortress while the other dies solvent-on-paper and broke-in-the-bank, purely because of the order in which cash moves through them.
Founders obsess over the P&L and barely look at this. They pick a model for its margins or its story and inherit its cash-flow shape without noticing they made a decision at all. Then, eighteen months later, growth — the thing they were told to want — starts strangling them, and they have no framework for why. The framework is that every model has a temperament: a fixed relationship between the moment you spend and the moment you collect. You don't get to renegotiate it after the fact. You live with it.
The income statement hides the thing that kills you
Profit and cash are measured differently on purpose, and the gap between them is where companies die. Profit is a claim about a period: over the last twelve months, revenue exceeded costs. Cash is a fact about a moment: right now, this much is in the account. A business can be relentlessly profitable across every period and still hit zero on a Tuesday, because profit can be locked inside inventory you've bought and receivables customers haven't paid.
The bridge between the two is working capital — the cash trapped in the gap between paying for something and getting paid for it. The cruel part is that this gap scales with growth. The faster you grow, the more cash the gap swallows. So the healthier the income statement looks, the deeper the hole can get.
Put numbers on it. Take a distributor doing $10M in revenue at a genuinely healthy 20% net margin — $2M of profit. Suppose working capital runs at 30% of revenue, normal for a business that holds stock and sells on terms. Now grow revenue 50%, to $15M. The extra $5M of revenue demands an extra 30% × $5M = $1.5M of cash locked into inventory and receivables before customers pay. You earned roughly $2M, and $1.5M of it never reached the bank — it's sitting in a warehouse and in other people's accounts payable. Grow 100% instead, and the incremental working capital is 30% × $10M = $3M, more than the entire year's profit. You are more profitable than ever and you have less cash than you started with. One supplier tightening terms, one bank pulling a line, and the profitable company is dead.
That is why "we're profitable, we're fine" is one of the most dangerous sentences a founder can say. It's an answer to the wrong question.
The archetypes, sorted by timing
Every business model collapses into a handful of cash-flow shapes once you sort them on one axis: do you collect before or after you spend?
| Shape | Timing | What it does to growth | What kills you |
|---|---|---|---|
| Negative working capital | Customer pays before you pay suppliers | Growth generates cash | Complacency; you rarely feel pressure |
| Positive working capital | You pay upfront, collect later | Growth consumes cash | Solvency — you starve while profitable |
| Lumpy / project | Big irregular inflows, steady outflows | Feast then famine | The gap between projects |
| Subscription | Pay to acquire now, collect slowly | Slow to build, then self-funding | Running out before payback |
Negative working capital is the best shape there is, and almost nobody designs for it deliberately. The customer's money arrives before the supplier's bill is due, so every new sale hands you float. Grow, and you accumulate cash instead of burning it. This is the structural magic under Amazon's early retail engine and under any business billed annually upfront while its costs are paid monthly. Sell a $600 annual subscription today against $360 of cost spread over the next twelve months, and each customer isn't just profitable — they hand you $600 of working capital on signup that you deploy for a year. Scale that and growth funds itself.
The inverse shape is the quiet killer, and it's the one the distributor above is trapped in. You buy or build first, sell second, collect third, often on net-30 or net-60 terms. Every unit of growth widens the gap between cash out and cash in. Hardware, wholesale, agencies staffing against net-60 invoices, anything with inventory — profitable businesses that discover, usually during their best year, that success is a cash emergency.
Lumpy cash flow is the consulting and construction shape: large, irregular inflows against steady monthly outflows. On average the numbers work. You don't pay rent and payroll on average, though; you pay them on the 1st. The feast-famine oscillation demands a reserve sized to the longest plausible gap between checks, and the failure mode is treating a fat quarter as the new baseline right before a dry one.
Subscription is negative working capital's slow cousin — but only if you survive the beginning. You pay the full acquisition cost today and recover it over many months. If CAC is $600 and you earn $50 of gross margin per customer per month, payback is twelve months. Acquire a hundred customers a month and you're spending $60,000 monthly to buy revenue you won't fully recoup for a year. Growth digs the hole faster: the harder you scale acquisition, the deeper you go before a cohort turns cash-positive. The model that becomes a cash machine at scale is a cash furnace at the start, and plenty of good subscription businesses die in the furnace.
Cash-on-delivery is a punishing shape most builders never have to imagine
I learned to see cash flow as timing rather than margin because I build for a market where the timing is brutal by default. Kommerce operates in cash-on-delivery commerce — the dominant model across low-trust emerging markets, where customers won't pay online for goods they haven't touched. In a Western prepaid model, money is captured the instant the order is placed; the whole cash-flow question is settled before a single box moves. COD inverts that completely.
In COD, you ship first. You front the product cost and the outbound logistics, the package travels for days, and only maybe the customer pays on delivery. The order isn't revenue when placed — it's an expense with a hope attached. And the shape has leaks a prepaid model never bleeds from. Failed deliveries: the courier can't reach the customer, and you eat the round-trip shipping on a sale that never happened. Returns at the door: the customer changes their mind on the doorstep, and now you've paid to send the product out and paid to bring it back, the goods possibly damaged in transit. Every one of these is cash that left and never came home.
This shape forces operational choices a prepaid business can treat as optional. You cannot survive COD without order verification — calling to confirm intent before you spend money shipping — because an unverified order is a lottery ticket you paid for. You push toward partial deposits to get skin in the game before you front the cost. You obsess over logistics control and delivery-success rates, because your entire cash conversion cycle runs through the courier's hands. None of that is optimization; it's the price of admission the cash-flow shape charges. A founder who copied a Stripe-checkout playbook into a COD market would be operationally naked and wouldn't understand why they kept running out of money on rising revenue.
Draw the timeline before you commit
The move is simple, unglamorous, and almost nobody does it before it's too late: before you commit to a model, draw its cash-flow timeline. Not a P&L projection — a timeline. Dates on the x-axis. For a single representative unit of sale, mark every moment cash leaves (buy inventory, pay for ads, ship the box, run payroll) and every moment cash arrives (deposit, delivery collection, invoice payment, subscription charge). Then do it again for a growth month, stacking the units, and watch the running cash balance as volume rises.
That picture answers the only question that matters: as I grow, does this shape feed me or eat me? If it eats you, you have two honest options. Redesign the terms so it feeds you — shift to upfront or annual billing, take deposits, stretch supplier payables, factor receivables, restructure delivery so collection happens sooner. Each lever drags your cash conversion cycle toward negative, and this terms-level work is the highest-return financial engineering most founders never touch, because it isn't as fun as the product. Or, if the shape is genuinely fixed by the market you serve, budget honestly for the cash growth will consume and decide, eyes open, where that cash comes from.
The shape also decides whether you're free. A cash-generative model funds its own growth and keeps the option to never raise a dollar. A cash-consuming one may force you to raise purely to fill the working-capital hole — a very different reason to sell equity than betting on a venture-scale outcome, and one that can put you across the table from investors whose fund math your business will never satisfy. The cash-flow shape you inherit when you pick the model partly writes your cap table before you've met a single VC.
Treat this the way you'd treat your data model. When money moves is as load-bearing and as expensive to change later as the schema your whole product silently agrees to depend on — a structural commitment made early, cheaply, and usually without the scrutiny it deserves, that then quietly governs everything downstream. You can rewrite your marketing in a week. You cannot easily rewrite the fact that you pay in January and collect in April.
You don't choose a business model and then figure out the cash flow. You choose the cash flow, and the model is just the story you tell about it.