Third Set Economics

RacketFuel Hungary · batch one · 1,000 boxes

What a customer costs and what a customer is worth

The blended figure in the GTM plan is 5,557 Ft a box. That number is flattered by club wholesale. This model separates the channels, builds LTV from subscription behaviour rather than assertion, and stress-tests the ratio. Every input below is live.

Assumptions

Drag anything. The whole page recalculates. Defaults are the route-A plan.

Per serving:
boxes per new customer
Benchmark ~32% when shown as the default option
Supplements run 5 to 8%
months. Stops the geometric tail running to infinity.

Unit economics, one box at a time

Net of VAT, COGS, WhiteFish pick and pack, packaging, courier, storage, card fees and a 4% failed-parcel allowance. Refused parcels return to stock, so only the shipping legs are lost, not the goods.

Contribution, first order
Contribution per box
first order, blended basket
Reorder at full price
one box, card, no acquisition cost
Subscription reorder

Where the first order's money goes

Contribution by acquisition route

The same box sold three different ways. Pick the route, move its lever, and watch what is actually left after you have paid to get the customer.

Net revenue per box
Contribution before acquisition
after VAT, COGS, fulfilment, courier
Acquisition cost per box
Contribution after acquisition

Where a box sold this way ends up

Route Lever Net rev / box CM before acq Acq / box CM after acq Margin on gross Whole batch this way

CAC, split honestly

One blended number hides two very different businesses.

Channel Boxes Acquisition spend Cost per box Cost per customer
The number that matters is 10,954 Ft, not 5,557 Ft. Club wholesale costs 2,938 Ft a box because 400 boxes move through 20 relationships, and it drags the blended average down by a third. Every DTC customer you acquire actually costs about . Underwrite the plan on that figure, not on the blend.

LTV, built rather than asserted

A box is roughly six weeks of supply at three sessions a week. Subscribers churn geometrically; everyone else reorders a fixed number of times. Capped at the horizon you set above.

Scenario Sub take Churn Sub reorders Orders / customer Customer life LTV

The ratio, stress-tested

Rows are what a customer costs, columns are what they turn out to be worth. Payback is how many orders it takes to get the acquisition cost back.

3.0x or better, healthy 2.0 to 3.0x, workable 1.0 to 2.0x, thin below 1.0x, loses money

What batch one is actually worth

The profit line is the part everyone looks at. The reorder pool is the part that decides whether there is a batch two.

Batch-one net profit
after fixed costs, HIPA and 9% TAO
Reorder pool, base case
Pool as a multiple of profit
contribution, zero extra acquisition cost
Retention scenario Reorders per customer Reorder pool, 400 customers vs batch-one profit
This is the whole argument for the subscription mechanics. The reorder pool carries no acquisition cost at all, so it converts almost entirely to profit. On the base case it is worth more than the entire batch-one profit line, from the same 400 customers, without spending another forint on marketing. Which is why the two numbers to watch from week 6 are subscription take-rate at checkout and cohort reorder rate by acquisition channel. They are the only two inputs on this page you can actually move.

What this model does not know

Every retention figure here is a hypothesis. There is no Hungarian subscription-churn data for sports nutrition, and no authoritative Hungarian e-commerce conversion rate exists at all. The 32% take-rate and the 5 to 8% churn band are international DTC benchmarks. Batch one exists to replace them with measurements.
Constant churn overstates the tail. The geometric model assumes a subscriber who has stayed nine months is as likely to cancel as one in month two. Real cohorts flatten, which helps, but early churn is usually worse than the average, which hurts more. The horizon cap is there so the number stays defensible. Leave it at 24 months when you show this to anyone.
Club LTV is missing and it is not small. A club that converts to a standing order of 15 boxes a month is worth far more than the 58,750 Ft it cost to sign, but its economics are a wholesale relationship, not a consumer subscription, so modelling it here would flatter the DTC numbers. Track it separately.
The first-order basket is a planning assumption. At 1.5 boxes it carries the model. If real baskets come in at 1.0, first-order contribution drops by a third and the plan-CAC column moves from comfortable to tight.
Cost base: WhiteFish Kft. 2026 fulfilment and courier quotes, including storage at 47 Ft a box and a 4% failed-parcel allowance. Benchmarks: subscription take-rate and churn from DTC industry data; Hungarian Meta CPC ~310 Ft from Shopify Hungary 2026. Route-A channel plan: 400 boxes club wholesale at 32% club margin, 480 affiliate, 120 event sampling, paid media reduced to brand defence. Figures are contribution, before fixed costs, unless a line says otherwise.