How Your Company Makes Money

The idea in one minute#

A surprising number of product managers cannot describe, in two minutes and with numbers, how their own company turns product into cash. Not approximately, specifically: who pays, what they are charged for, what it costs to earn that, and what is left. Without that description every prioritisation argument is conducted on intuition about value, and intuition about value is exactly what the description replaces. It is also the fastest way to find out that your product’s revenue comes from somewhere other than where your roadmap is pointed, which is common enough to be worth checking before anything else in this topic.

A picture#

flowchart LR
  CUST[":i-users: <b>Who pays</b>"] --> UNIT{":i-tag: <b>What they are charged for</b><br/><small>the value metric</small>"}
  UNIT --> REV[":i-banknote: <b>Revenue</b>"]
  REV --> COGS[":i-receipt: <b>Cost of revenue</b><br/><small>hosting, support, inference, payments</small>"]
  COGS --> GM[(":i-percent: <b>Gross profit</b>")]
  GM --> OPEX[":i-building-2: <b>Everything else</b><br/><small>salaries, sales, marketing</small>"]
  OPEX --> PROFIT[(":i-coins: <b>What is left</b>")]
  ACQ[":i-megaphone: <b>Cost to acquire</b>"] -.->|"paid up front, earned back over time"| GM
  class CUST neutral
  class UNIT queue
  class REV,OPEX compute
  class GM,PROFIT memory
  class ACQ io

Five boxes and one arrow that crosses time. The acquisition cost is spent now and recovered later, which is where most of the interesting arithmetic lives.

How it really works#

The six questions#

Answer these about your own company. If you cannot, the rest of this topic is theoretical.

1. Who pays? The named role or segment, not “our customers”.
2. What are they charged for? Seats, usage, transactions, a flat fee, a share of something.
3. How much, on average? The real average including discounts, not the list price.
4. What does it cost to serve them? Hosting, support, payment fees, inference, third party licences.
5. What does it cost to acquire them? Sales and marketing spend divided by customers won.
6. How long do they stay? Which decides whether question five was affordable.

Six numbers. Most companies will give them to you if you ask, and the asking is itself useful, because the person who owns each number is the person to talk to when a product decision moves it.

Revenue is not one thing#

Most companies have several revenue lines with very different economics: a core subscription, an expansion line, perhaps services, perhaps payments or a marketplace take. They carry different margins and grow at different rates, and the one that looks largest is often not the one that matters at the margin.

The question to ask is where the next dollar comes from, which is frequently a different line from where the last hundred came from. A roadmap pointed at the historical line while the business grows on another is a common and quiet misalignment.

Where product actually moves the number#

Product has leverage on four of the six: what they are charged for, how much, what it costs to serve, and how long they stay. It has indirect influence on acquisition cost, through conversion and through what the product makes provable in a sales conversation.

That list is worth internalising because it tells you which business arguments you can make credibly. “This improves retention, which extends payback” is a product argument. “This reduces sales cost” usually is not, unless you can name the mechanism.

The timing problem#

Acquisition is paid now and earned back over months. That single mismatch generates most of the financial vocabulary in this topic: payback period, runway, burn, capital efficiency. All of them are different ways of asking whether you can afford the gap between spending and earning.

What fills the box in 2026#

The numbers live in the billing system, the warehouse and the finance team’s model, and the three frequently disagree for the reasons in Working With Data. Ask finance for their version and use that one in any conversation with them, while keeping your own for product reasoning. Arguing about whose number is right is almost never the valuable conversation.

Sector lens#

Remember this#

• Six questions: who pays, for what, how much, cost to serve, cost to acquire, how long they stay.
• Revenue is several lines with different economics. Ask where the next dollar comes from.
• Product moves four of the six directly. Know which arguments you can make credibly.
• Acquisition is paid now and earned later, and that mismatch generates the rest of the vocabulary.
• Use finance’s numbers when talking to finance. Arguing about whose figure is right rarely helps.

Try it#

1. Write the six answers for your company from memory. Then check each one and note how far off you were.
2. Ask finance which revenue line is growing fastest. Compare that to where your roadmap points.
3. For your last three roadmap items, name which of the six numbers each one moves. Any that move none are worth questioning.

Check yourself#

1. Why does the gap between the historical revenue line and the marginal one matter to a roadmap?
2. Which of the six numbers can product move directly, and which can it only influence?
3. Why does the timing mismatch between acquisition and earning generate so much financial vocabulary?