Does the model make money per order? Marketplace unit economics before scale

1. Before you start

A marketplace does not buy and resell — it matches a buyer with a seller and keeps a cut of the deal. That cut is the take-rate: the share of the transaction value the marketplace keeps. If a buyer pays $50 for something and the marketplace keeps 15%, its revenue on that deal is $7.50, and the seller nets $42.50. Unit economics asks a blunt question about that $7.50: after you pay everything one transaction causes — payment fees, the delivery, support — does the marketplace keep anything, or does the deal cost more to run than it earns?

Three honest statements before you start:

This is a Decide course. You read a situation, learn the idea, and make a call. You do not write or run any code.
The company below, Basketful, and the company in the final scenario, Draftly, are composite — invented from ordinary, realistic figures so the arithmetic is clean. No number is a claim about any real company.
This is not a certification. It proves, to you, that you can defend a call on whether a business model earns money at the level of one transaction.

You need only arithmetic. The hard part is judgment: knowing which costs a single deal really causes, and refusing the comforting story that scale fixes everything.

2. The Situation

Basketful is a grocery-delivery marketplace: shoppers order from independent neighbourhood grocers through the app, a courier delivers, and Basketful keeps a commission. Orders are up 140% this year, the board is thrilled, and the founder wants to triple the marketing budget to “get to scale, where the economics finally work.” The head of finance has gone quiet, because she has run one number the growth deck skips: what a single order actually earns after the courier, the card fee, and support are paid. You have to decide whether pouring money into growth makes this model bigger, or just makes the hole deeper — and one figure on that page decides which.

3. What you’ll be able to do

After this course you will be able to:

Compute a marketplace’s contribution per transaction — take-rate times order value, minus the variable cost the order causes — and say whether a single deal makes money at all.
Judge the claim “we’ll fix the economics with volume” and say exactly when it is true and when scaling only multiplies the loss.
Bring in customer acquisition cost and decide whether the lifetime contribution of a customer ever repays what it cost to win them.
Name the one lever — take-rate, variable cost, frequency, or acquisition cost — that flips a model from value-destroying to viable, and defend the choice.

4. Prerequisites & time box

Prerequisites: arithmetic, and the idea of a variable cost (a cost caused by one more unit of activity) versus a fixed cost (one you pay regardless). Both are defined again as they come up. No spreadsheet, no finance background. If you have met contribution margin before, you are ahead; if not, section 6 builds it from scratch. New to the Decide hall? Read its short how-to-read page first.

Time box: about 18 minutes of reading (measured), plus your own thinking time on the call in section 7. That is under the 25-minute cap for a concept course.

Difficulty: 5 / 8. This sits a notch above a first margin decision. You are not just computing one number and reading it off; you weigh three interacting levers — the per-deal contribution, the cost to acquire a customer, and how often that customer comes back — and a plausible, confident wrong answer (“scale will fix it”) is waiting at every step. There is real arithmetic, but the trap is judgment: knowing which number actually governs the decision. A version with the figures handed to you and a single lever to pull would be a 3; the weighing of levers against a seductive false story is what puts it at 5.

5. The case & where the numbers come from

Basketful is a composite marketplace: a grocery-delivery model assembled from the ordinary economics a two-sided, delivery-based marketplace would recognise — a modest take-rate, a courier cost that dominates the variable line, thin per-order margins. The figures are chosen for clean arithmetic and are not drawn from or claimed about any real company. The definitions — take-rate, contribution, variable cost, customer acquisition cost, lifetime value — are standard and cited in section 11. Every dollar in the table below is an in-course assumption; every later number is computed from these.

Item	Figure
Average order value (GMV per order)	$45
Take-rate (commission Basketful keeps)	11%
Variable cost per order (see breakdown in section 6)	$6.00
Customer acquisition cost (CAC)	$22
Orders per customer per year	10
Average customer lifespan	2 years

The variable cost per order of $6.00 breaks down as: courier payout of $3.80 (this is already net of the small delivery fee the customer pays, which is folded in here rather than counted as separate revenue); payment processing, $1.30; support, insurance and packaging, $0.90. Each is a cost a single order causes; none is Basketful’s office lease or salaried staff, which are fixed and sit outside this decision.

6. The Concepts

Contribution per transaction

Contribution per transaction is what one completed deal keeps after the costs that deal causes. For a marketplace it is: take-rate times order value, minus variable cost per order.

Basketful’s revenue on one order is its take-rate times the order value: 11% of $45 = $4.95. That is the only money Basketful keeps; the other $40.05 belongs to the grocer. Against that $4.95 sits the $6.00 the order costs to run. So:

Contribution per order = $4.95 revenue − $6.00 variable cost = −$1.05.

Every single order loses Basketful $1.05 before a cent of office rent or engineering salary is paid. This is the number the growth deck skipped. Note what it is not: it is not gross margin on the $45 (Basketful never owned the groceries), and it is not the courier cost alone. It is the marketplace’s own cut minus the marketplace’s own per-order costs — and here that is negative.

(An interactive calculator sits here — enter your own order value, take-rate, variable cost, acquisition cost and order frequency, and it returns the contribution per order, the lifetime value of a customer, and whether the model covers what it cost to win that customer. Change the take-rate or the acquisition cost and watch the decision flip.)

The volume trap

The founder’s instinct — “get to scale, where the economics work” — is right for fixed costs and wrong for variable ones, and confusing the two is the trap. Fixed costs (the lease, salaried staff) are paid once and spread thinner over more orders, so scale genuinely helps them. But contribution per order is built only from variable costs, which recur on every single order. Volume does not dilute them; it repeats them.

Watch what scale does to Basketful’s −$1.05 per order:

At 100,000 orders a month: 100,000 × (−$1.05) = −$105,000 of contribution a month.
Double the orders to 200,000: 200,000 × (−$1.05) = −$210,000 a month.

Doubling volume doubled the loss. Spending more to acquire more customers, at negative contribution, buys a bigger negative number. “We’ll fix it with volume” is only true once contribution per order is positive — then each extra order adds margin and scale spreads the fixed costs on top. Below zero, growth is the accelerator on the wrong car. The rule: fix the per-unit contribution first; scale second. Basketful’s real question is not “how do we grow?” but “what makes one order profitable?” — raise the take-rate, or cut the variable cost. Lift the take-rate to 16%: revenue becomes 16% of $45 = $7.20, contribution becomes $7.20 − $6.00 = +$1.20 an order. Now, and only now, does volume help.

Customer acquisition cost and payback

A positive contribution per order is necessary but not sufficient. You still paid to win the customer, and that cost has to come back. Customer acquisition cost (CAC) is the fully loaded cost of acquiring one paying customer — paid marketing plus the first-order promo credit — here $22.

You recover CAC out of the contribution each customer throws off over their life, not out of one order. Two numbers turn a per-order figure into a per-customer one:

Lifetime orders = orders per year × lifespan = 10 × 2 = 20 orders.
Lifetime contribution (LTV) = contribution per order × lifetime orders.

At the fixed-up take-rate of 16% (contribution $1.20/order): LTV = $1.20 × 20 = $24. Compare that to the $22 it cost to acquire the customer: $24 − $22 = +$2. The model finally clears the bar, but barely — a customer is worth $2 more than they cost. The payback in orders is CAC ÷ contribution per order = $22 ÷ $1.20 ≈ 18 orders, about 1.8 years at 10 orders a year — just inside the 2-year lifespan. Push CAC up to $30 (a pricier ad market) and the customer now costs more than the $24 they return: back below the line. Cut CAC to $12 and the margin of safety opens up. So even with contribution fixed, the call still turns on acquisition cost and how often the customer comes back. Two layers must both hold: each order contributes, and a customer’s lifetime contribution repays the cost to win them.

7. Your Call

You have seen how contribution per transaction, the volume trap, and CAC payback decide Basketful’s call. Now a different one lands on your desk.

Draftly is a marketplace that matches small businesses with freelance designers for one-off logo and brand-identity projects. The average project is $300, Draftly keeps a 15% take-rate, and the variable cost it causes per project — payment processing, escrow and dispute handling, support — is $18. A business commissions about 1.5 projects a year and stays for 2 years. Draftly’s CAC is $95. A competitor has just launched at a 10% take-rate, so raising Draftly’s commission risks driving designers and buyers to the cheaper venue. Your job is to tell the founder whether this model is viable as it stands, and if not, which lever to pull.

How this differs from the taught case (the transfer): this is a different sector and company (a freelance-services marketplace, not grocery delivery), the figures are different so the arithmetic must be redone rather than recalled, it asks a different kind of decision (diagnose which lever fixes the model, not a single accept/reject), and it adds a constraint the Basketful case never had — a competitor’s take-rate that caps how far you can raise your own. The core concept is the same: whether the business model earns money at the level of one transaction, before you spend to scale it.

8. Self-check

Before you write the memo, make sure you can say each of these in one line:

What is the difference between contribution per transaction and the marketplace’s revenue per transaction, and why does only one of them decide whether growth helps?
When is “we’ll fix it with volume” true, and when is it the accelerator on the wrong car?
Contribution per order can be positive while the model still loses money per customer — what second test has to pass, and what number would flip it?

If any is fuzzy, reread section 6 — contribution per transaction, the volume trap, and CAC payback are the whole course.

9. Stretch

Push the decision further on your own:

Back at Basketful: at the 16% take-rate the model clears CAC by only $2 a customer. What single 10% swing hurts more — a 10% rise in courier cost, or a 10% rise in CAC? Work both through and say which lever the founder should guard most.
Draftly is weighing a $40,000-a-year “concierge” team that lifts repeat frequency from 1.5 to 2.2 projects a year. At what number of active customers does that fixed cost pay for itself — and how does that change the make-or-break lever?
The genuinely hard one: a rival marketplace runs at −$0.50 contribution per order on purpose, funded by investors, to take the market before fixing economics. Write the one paragraph you would give your board on when that is a strategy and when it is a slow bankruptcy — and what evidence would tell the two apart.

10. Ship it — your decision memo

Write a one-page memo to Draftly’s founder. State the call (the model is not viable as it stands: $81 of lifetime contribution against a $95 CAC, so each acquired customer loses about $14). Show the two-line arithmetic (contribution $300 × 15% − $18 = $27 per project; lifetime $27 × 3 projects = $81 < $95 CAC). Name what you rejected and why — accepting it on the strength of a positive per-project figure, and raising Draftly’s take-rate further when a rival already undercuts it at 10%. Name the lever you would pull (repeat frequency or CAC) and the number that would change your mind (lifetime contribution rising above $95). Keep it to a single page a founder grasps in two minutes. This memo is your own argued claim — something you can defend in a room, not a credential.

11. Sources

Basketful and Draftly, and every dollar figure attached to them, are composite — invented from ordinary, realistic figures for clean teaching arithmetic, not drawn from or claimed about any real company. The concept definitions used to reason about them are standard; references below.

Concept / claim	Source (publisher)	URL	Accessed
Contribution margin = revenue − variable cost	Wikipedia — Contribution margin	https://en.wikipedia.org/wiki/Contribution_margin	2026-07-20
Definition of a variable cost	Wikipedia — Variable cost	https://en.wikipedia.org/wiki/Variable_cost	2026-07-20
Take-rate and marketplace/GMV terms	Wikipedia — Online marketplace	https://en.wikipedia.org/wiki/Online_marketplace	2026-07-20
Gross merchandise volume (order value base)	Wikipedia — Gross merchandise volume	https://en.wikipedia.org/wiki/Gross_merchandise_volume	2026-07-20
Customer acquisition cost	Wikipedia — Customer acquisition cost	https://en.wikipedia.org/wiki/Customer_acquisition_cost	2026-07-20
Customer lifetime value (LTV)	Wikipedia — Customer lifetime value	https://en.wikipedia.org/wiki/Customer_lifetime_value	2026-07-20

Next up

Finished this call? Continue the Strategy track:

Where’s the moat? Telling a durable advantage from a good quarter → · Browse all courses