Recurring, capacity, or funnel: which revenue engine are you?

Ask most owners how they forecast revenue and the honest answer is: last year plus a bit. It is not stupid — it is often roughly right — but it is useless the moment you want to changesomething, because a growth percentage has no parts. You cannot pull a lever on 8%.

A revenue model has parts. And here is the useful simplification: almost every business runs on one of three engines. Once you know which one you are, you know exactly which handful of numbers drives your top line — and which ones are worth measuring every month.

Engine one: recurring

You sell access that renews. Software subscriptions, memberships, monthly retainers, subscription boxes, maintenance contracts. Revenue this month is mostly revenue last month, which is the great gift of the model — and the trap, because decline is just as sticky as growth.

Your revenue is a balance, not an event. It moves through four drivers: how many new customers you add, what they pay on average, how many you lose, and how much existing customers expand. Everything else is commentary.

The number to watch is not MRR. It is churn as a share of opening balance, because that single figure sets the ceiling on how big you can get: at 6% monthly churn, you top out at roughly sixteen times whatever you add each month, no matter how good sales gets.

Engine two: capacity

You sell time or space, and you have a hard ceiling. Salons, restaurants, hotels, clinics, trucking fleets, agencies billing hours. There are only so many chairs, rooms, trucks, or hours in a week, and any one that goes unsold is gone forever.

Revenue is a multiplication: units × sellable slots × utilization × price, plus whatever you sell alongside it. Four chairs, forty slots a week each, 68% filled, $65 a visit. That is your business, and each of those four terms is a different lever with a different cost to pull.

Utilization is almost always the cheapest lever and the last one people look at. Moving 68% to 76% adds 12% to revenue with no new staff, no new space, and no price increase — it is pure yield. Adding a fifth chair adds cost immediately and revenue only if you can fill it.

Engine three: funnel

You buy attention and convert it. Ecommerce, DTC, marketplaces, lead generation, anything where paid acquisition is the primary growth motor. Revenue is the output of a chain, and every stage multiplies: spend → sessions → conversion → orders → average order value, adjusted for returns and repeat purchases.

The dangerous property of a funnel business is that it looks like it scales linearly and does not. Doubling spend rarely doubles orders, because cost per click rises as you push past your best audiences. A model that assumes constant CAC will overstate next year badly.

• Conversion rate is the highest- leverage term, because it multiplies everything downstream at zero marginal cost.
• Repeat rateis what decides whether you can outbid competitors — it is the difference between affording one purchase of CAC and three.
• Returns quietly eat the margin you use to fund acquisition, and almost nobody models them.

What if you are more than one?

Plenty of businesses are. A gym with memberships (recurring) and personal training slots (capacity). An agency on retainers (recurring) plus project work (capacity). A DTC brand with one-time orders (funnel) and a subscribe-and-save tier (recurring).

Model them separately and add them up. Do not average them together — a blended forecast hides the fact that the two halves respond to completely different actions, and it is precisely the mix between them that you most want to see moving.

Why the shape matters more than the forecast

The forecast will be wrong. That is fine and expected. What a driver-based model buys you is not accuracy — it is diagnosis. When revenue comes in 12% light, a growth-percentage model tells you revenue was 12% light. A driver model tells you utilization held but the average ticket fell, or that new customers were on plan and churn spiked.

One of those is a number. The other is a decision.