Every pricing model is a loan. Bill hourly or on completion and you fund your client's operations until they pay. Bill a retainer or an annual subscription in advance and your customers fund yours. Three variables set the direction: when cash arrives relative to when you incur the cost to deliver, whether delivery cost is fixed or scales with each customer, and how much revenue predictability you need to plan against committed costs. Answer those and the model mostly picks itself. Ignore them and you can run healthy revenue at a healthy margin and still miss payroll.
Search this question and you get taxonomy. Ten models, a paragraph each, and an instruction to survey your customers and find where they see value. None of it tells you whether you can make payroll. Perceived value sets what number goes on the page; it says nothing about when that number becomes money in your account.
Take two firms at $1.2M of revenue and 45% contribution margin, identical on every line anyone looks at. Firm A bills hourly, invoices at month end, offers net 30, and collects on a 52-day DSO. Firm B bills a monthly retainer on the first for the month ahead and is paid within five days.
Firm A carries about $171,000 in receivables. Firm B carries about $16,000, and is also sitting on a month of customer money for work it hasn't delivered, about $50,000 on average, a liability on paper and cash in the bank in practice. Count it and the gap is $205,000, 17% of revenue, between two businesses that look identical in every report they produce. Neither owner finds this in their P&L, which isn't built to tell them.
Same revenue, same margin, opposite sides of the loan. Nothing but the pricing model produced that.
Five structures govern how a customer buys from you. Value-based is usually listed as a sixth; it isn't one, and the section after the table explains why. Every scenario below is illustrative.
How it works: You sell units of time and bill for the ones you use.
Cash timing: The worst of any model, and the loan always runs away from you. You pay labor as it's consumed and collect two to three months after the work is done.
Margin shape: Capped by revenue per head. The one real lever is delivery mix: bill at $250 against senior cost of $150 and you run 40% gross margin; push 70% of the work to staff costing $70 and the same rate returns better than 60%. That pyramid is the economic model of every law and consulting firm. What it can't do is grow revenue without growing people.
What breaks it: A $25,000/month engagement costing $14,000/month to deliver. Invoice at the end of month one, collect on a 52-day DSO, and the first dollar lands around day 82 — by which point you've funded $37,800 of payroll yourself. The engagement contributes $11,000 a month, so you carry three and a half months of its own contribution before it pays you anything. If delivery cost is mostly senior labor and engagements run past 60 days, hourly is financing your client's business.
Who it fits: Truly unpredictable scope, or short work where the client insists on auditing inputs. A reasonable way to start and a bad place to stay.
How it works: One price for a defined deliverable.
Cash timing: Set by your milestone schedule, which most founders treat as a contract detail rather than a decision about how much to lend. Billed on completion, it's the largest loan you'll ever make a client.
Margin shape: Improves if you get faster than you priced it, and improves more if you can deliver the same scope with cheaper labor. Templating and reusable work count here; heroics don't.
What breaks it: A $60,000 project, 90 days, $48,000 to deliver, billed on completion and collected 30 days later. You don't spend that $48,000 on day one, so the balance you finance averages about $30,000 over the 120 days. At 14% — closer to a real working capital line than the 10% textbooks use — that's $1,380 of carry against $12,000 of contribution. 11.5% of it, gone before anyone scopes a change order. A 90-day project billed on completion at 20% margin is a 90-day interest-free loan to your client.
Who it fits: Repeatable deliverables where you have history on actual delivery hours. Deposits and milestones fix most of the cash problem; nothing fixes pricing a project you've never done.
How it works: A recurring fee for ongoing access or a standing scope.
Cash timing: The first model where the loan runs toward you by default, provided you bill in advance. Cash arrives before the cost is incurred, which is the best a services business can do.
Margin shape: Stable on paper and erosive in practice, because revenue is fixed and hours aren't.
What breaks it: An $8,000 retainer scoped at 40 hours a month is a $200 effective rate. Let it drift to 58 hours, which happens gradually and without a conversation, and the rate is $138. That's a 31% cut to the rate, and margin runs ahead of it: at $100 an hour loaded, contribution falls from $4,000 to $2,200, a 45% cut, with no change to revenue and no signal in your financials. Retainers fail silently, which is why they need a monthly hours-against-scope review that project work doesn't.
Who it fits: Ongoing work with a scope you can define, at a firm with enough committed fixed cost to need predictable revenue.
How it works: Recurring fee for continued access, usually per seat or per tier.
Cash timing: The best of any model. Annual prepay is the biggest loan your customers will ever make you: negative working capital, their money funding your growth.
Margin shape: The only structure with gross-margin expansion built into it, and only if delivery cost is truly fixed. If a human is in the loop on every account, you have a services business with subscription billing, and the forecast leverage doesn't exist.
What breaks it: Spending the float. Annual prepay hands you twelve months of cash against twelve months of delivery you haven't done. Spend it in five and you're funding this year's delivery out of next year's bookings, which works exactly as long as bookings grow. They don't have to shrink to hurt you, only flatten: zero churn and flat bookings still produces a cash cliff, because the float stopped growing and the obligation didn't. It's the cheapest money you'll ever raise and the only lender that never asks for it back, but every dollar is spoken for. Watch it against your CAC payback, not your bank balance.
Who it fits: Products where one more customer costs near nothing and the value is continuous rather than episodic.
How it works: The customer pays for what they consume.
Cash timing: In arrears by definition: deliver, meter, then bill. You're lending for the length of the metering period. A small loan, but always yours to make.
Margin shape: Good, at the gross line. Variable costs track revenue almost automatically, so gross margin holds as you scale. Operating costs track nothing at all.
What breaks it: You've taken on your customer's volatility. Usage that swings 40% either way turns a $9,000 month into a $5,400–$12,600 range. Ten customers help less than you'd hope: independent swings would damp the wobble to about 13% either way, not to zero, and they aren't independent, because usage tracks your customers' own seasonality and they share a calendar. Correlated, you keep the full 40%. That's a $72,000 monthly revenue spread, about $32,000 of contribution at a 45% margin, against an operating base that doesn't flex.
Who it fits: Businesses where consumption really does track value received, and where you can survive a slow quarter you didn't choose.
Before committing to any of these, run one deal through the pricing profitability calculator. Revenue, delivery cost and overhead in, real contribution out. No email required.
| Model | Cash timing — who lends | Revenue predictability | Margin as you scale | Collection risk (DSO) | Primary failure mode |
|---|---|---|---|---|---|
| Hourly | You lend. Cost first, cash 60–90 days later | Low | Capped by revenue per head; mix is the only lever | Long, but billed in small increments | Financing the client |
| Fixed-fee / project | You lend, on a schedule you set; largest billed on completion | Low, and lumpy | Improves on speed and on labor mix | Largest single exposure; deposits cut it | Scope creep against a fixed price |
| Retainer | They lend, if you bill in advance | High | Erodes silently | Low | Unpriced scope drift |
| Subscription | They lend. Annual prepay is negative working capital | Highest | Expands, if delivery cost is truly fixed | Very low | Spending the float |
| Usage-based | You lend for the metering period | Low to moderate | Gross margin holds; opex doesn't flex | Moderate | You absorb the customer's volatility |
| Value-based | Depends on the model you wrap it in | Depends | Highest ceiling | Depends on terms | No agreed number to price against |
Every cell in the last row says depends. That's the next section.
Value-based pricing is not the sixth option. The five structures above decide who finances whom. Value-based decides whether the relationship is worth financing at all, meaning how much they pay relative to what they get. They're different questions and you need an answer to both.
So value-based should be considered every time, including when you've already chosen one of the other five. A retainer can be value-based. So can a project fee, a subscription tier, or a usage rate. The packaging is just a digestible way to buy; what they pay across it should still sum to something defensible against the value you deliver.
Here's where our view departs from the standard advice, which is to capture more of the value you create. For an early-stage company we think that's backwards. Aim for the value the customer receives to be roughly 10x what they pay you.
That sounds like leaving money on the table. In the short run it is, and what it buys is the only durable pricing power there is. A customer paying $20,000 for something worth $200,000 to them won't run a competitive process at renewal or grind you on rate, and won't cancel in a bad quarter. You're not the line item they cut. You're the one they protect. That gap is your switching cost, and early on it's worth more than the margin it costs.
It's a starting posture, not a permanent one. As the value gets proven and referenceable you capture more of it. Compress too early and you never build the stickiness that made compression possible; never compress and you stay underpriced forever.
The discipline this imposes is simple, and most firms fail it: you have to be able to name the number. Estimating it doesn't count. You need a sentence the client will agree to out loud. "This takes 22 hours a month off your controller and stops the reforecast slipping two weeks" is a number. "Significant efficiency gains" is not. If you can't get the client to agree to a number, you aren't doing value-based pricing. You're doing fixed-fee with a longer sales cycle.
1. Is your delivery cost fixed or per-customer? If the tenth customer costs roughly what the first did, subscription is available to you and you should take it. If each one consumes real human hours, subscription billing on a services cost structure will show you leverage in the forecast that never arrives in the bank.
2. How much of your cost base is committed within 90 days? High committed cost demands predictable revenue, which means retainer or subscription rather than project work. Low committed cost means you can carry the lumpiness and chase larger, less frequent deals. Run your break-even before deciding which you are.
3. Retainer or project? Project pricing if the work has a defined end and you have history on delivery hours. Retainer if the work is truly continuous and you'll run the monthly scope review. Without it, a retainer is a project that never ends and never gets repriced.
4. Who carries which variance? There are two, and they move separately. Usage-based hands you the revenue variance. Retainer and subscription hand the customer the spend variance, since they pay the same whether they use you or not, and leave the delivery-cost variance with you, which is what the 40-to-58-hour drift is. Hourly protects your margin per hour and gives you all the revenue variance instead. No model lets you put down both, so pick the one your balance sheet can carry.
Whatever you land on, payment terms are a separate negotiation from price, and the place where an agreed number quietly becomes a different one.
Most writing on this stops at the recommendation. The transition is the part you live through, and four things break in it.
The revenue trough is a negotiation, not arithmetic. Going from arrears hourly to advance retainer, the arithmetic favors you: one month where you bill for the work you just finished and the month ahead, then straight into steady state. That's a permanent one-month pull-forward of cash, free. The catch is that almost no client accepts a double-charge month, so you waive the arrears invoice and the windfall becomes a one-month hole. Decide before you announce whether you'll ask for it, phase it over a quarter, or eat it. In the other direction the trough is real and can't be negotiated away: you bill nothing in the switch month, and every month after lands thirty days later.
Grandfathered customers become permanent. Every founder intends to migrate legacy accounts "next year." Two years on they're a third of revenue on terms nobody would sign today, and because they're your best references they're the hardest to reprice. Set the migration date when you set the model, and tell those customers then.
Comp plans stay priced on the old model. Pay salespeople on booked project value and move to subscription, and they'll keep selling projects, correctly, because that's what you pay them for. Tie delivery bonuses to billable hours and move to fixed-fee, and efficiency becomes a pay cut for the people you need to be efficient. The comp plan changes on the same day as the pricing model, or the pricing model doesn't change.
Your blended DSO will lie to you. Advance-billed revenue drags the average down fast, so the headline number improves while you run two billing motions through one AR process. Collections on the legacy arrears book slip during the overlap and the blend hides it. Track the old book on its own until it's gone.
The structure picks who finances whom, how predictable revenue is, and whether margin can improve with scale. All of that is ceiling. Whether you get near it is decided later, deal by deal, in discounting, scope and terms — a different problem with different arithmetic.
Run the model you're considering through the pricing profitability calculator and the unit economics calculator first. Both are free, and neither asks for your email. If the numbers you'd need don't yet exist in a form you trust, that's the actual first problem, and the one our Cash Clarity Sprint is built to solve.