A great idea with a broken business model is still a failure. Your business model is simply how value gets created, delivered, and—crucially—captured as revenue. Get this right early and everything downstream gets easier. Get it wrong and you’ll feel it in every cash-flow report.
This guide walks you through choosing a revenue model and sanity-checking whether the numbers actually work.
Free download: Milk Spider One-Page Business Plan (Word) — a living one-pager to capture your model, customer, numbers, and first milestones.
Pick how you’ll make money
Most businesses use one (sometimes two) of these revenue models. Here’s when each tends to fit:
- One-time sale — A product or project sold once. Simple and easy to understand, but you have to keep finding new customers to keep revenue flowing.
- Subscription / recurring — Customers pay on a repeating schedule. Predictable revenue and higher lifetime value, but you have to keep earning that renewal every cycle.
- Usage-based — Customers pay for what they consume. Aligns your revenue with the value delivered, and scales naturally with heavy users.
- Marketplace / commission — You connect buyers and sellers and take a cut. Powerful at scale, but hard to start because you need both sides at once.
- Freemium — A free tier brings people in; a paid tier captures the ones who need more. Great for reach, but only works if enough free users convert.
- Services / consulting — You sell your time and expertise. Fast to start and high-margin, but it’s capped by the hours you can work unless you productize it.
Choose the one that matches how your customer wants to buy and how often they’ll get value. A problem people face once a year rarely supports a subscription. A tool people rely on daily often does.
Understand your unit economics
This is the part founders most often skip—and most often regret skipping. Unit economics is just the money math on a single customer or sale. If one sale doesn’t make sense, a thousand won’t either. You need rough estimates for five numbers:
- Price — what you charge per unit, month, or transaction.
- Cost to deliver (COGS) — what it costs you to deliver that one unit.
- Gross margin — price minus cost, as a percentage. This is the money left to run the business.
- Customer acquisition cost (CAC) — what you spend on marketing and sales to win one customer.
- Lifetime value (LTV) — the total profit you earn from one customer over the whole relationship.
The single most important relationship here is LTV vs. CAC. If it costs you more to acquire a customer than that customer is ever worth, you don’t have a business—you have a leak. A common rule of thumb: you want lifetime value to be roughly 3× your acquisition cost or better, with the cost paid back within a reasonable window.
You won’t have perfect numbers at the start. Estimate honestly, label your assumptions, and update them as real data comes in.
Test your riskiest assumptions
Every business model rests on a few assumptions that must be true. Maybe it’s "customers will renew month after month," or "I can acquire customers for under $40," or "people will pay before they see results." List the five assumptions your model most depends on, then—for each—write the cheapest way to test it before you’ve sunk real money in.
This turns vague optimism ("I think this will work") into a concrete plan ("I’ll know whether this works after I test these three things").
Look for durable advantages
A model that makes money today is good. A model that’s hard to copy is better. As you choose, ask whether your business builds any of these over time:
- Switching costs — it gets harder for customers to leave the longer they stay.
- Network effects — the product gets more valuable as more people use it.
- Data or brand — you accumulate something competitors can’t easily replicate.
You don’t need all of these on day one. But knowing where your durability could come from helps you steer toward it.
FAQ
Can I use more than one revenue model?
Yes, and many businesses do—for example, a base subscription plus usage-based overage charges. Just don’t make it confusing for the customer. Start simple and layer complexity only when it clearly helps.
What if I don’t know my costs yet?
Estimate. Use the best numbers you can find, mark them as assumptions, and refine them once you have real sales. The point is to catch obviously broken math early, not to be precise to the penny.
My LTV:CAC looks bad. Should I quit?
Not necessarily—but it’s a signal to change something before you scale. Often you can fix it by raising prices, improving retention, or finding a cheaper acquisition channel. Scaling a broken model just loses money faster.
How do I estimate lifetime value with no customers yet?
Make a reasonable guess based on your price and how long you expect customers to stay, then treat it as a hypothesis to validate. Early real data will replace the guess quickly.