This article is also available in Danish.
Forretningsmodel

Business model: What it is, and how to change yours

By Daniel Wegener 18 August 2026 5 min read

A business model is the answer to who you serve, what you solve, how you deliver it, how you get paid, and what it costs you. It can change without the product changing, and that is usually where the largest gain sits for a company that has stalled.

Most owner-managers can explain what their company does. Fewer can explain how it makes money without using the word revenue.

That sounds like a quibble. It is not. The difference shows up the day revenue holds while margin slips, and nobody can point at why.

What a business model is

Five questions, and they should be answerable without anyone starting to explain:

This is not a strategy. A strategy is what you want to achieve. The business model is the machinery that has to produce it.

The distinction matters, because many strategy problems are really model problems. A company that cannot grow without hiring at the same rate does not have an ambition problem. It has a cost structure that scales linearly with revenue.

Draw it in half an hour

The Business Model Canvas is nine boxes: customer segments, value proposition, channels, customer relationships, revenue streams, key resources, key activities, key partners and cost structure.

Take a large sheet of paper and fill them in. One requirement: draw the model as it actually works, not as it appears in the material you send the bank.

The best version of the exercise is to have three or five managers draw it separately and then compare. If there is disagreement on the facts, meaning on who pays for what, you have found something more valuable than any strategy discussion would have given you.

When it needs to change

Not because three years have passed. There are five concrete triggers:

If none of the five is present, there are better places to spend a week.

Find the box carrying the risk

This is the step most often skipped, and it is why most changes land in the wrong place.

The nine boxes are not equally important. One of them usually carries the risk. Four questions find it: what happens if the largest customer leaves, do costs scale as fast as revenue, could a competitor copy the value proposition in six months, and do customers pay before or after you incur the cost.

There is rarely more than one place where the answer genuinely hurts. The method and a worked example are in stress-test your business model.

Choose an alternative that fits the problem

There is a limited number of ways to make money. Researchers at the University of St. Gallen mapped 55 recurring patterns, and the catalogue has since been extended with platform and circular-economy patterns among others.

The point is not the inspiration. The point is that each pattern has known ways of failing: subscription without a reason to stay, platform without a plan for the cold start, licensing without clarity on what is being licensed. The patterns and their pitfalls are covered in 66 business model patterns.

Choose the pattern that solves the box carrying the risk. Subscription is not better than hourly billing. It is better if the problem is unpredictable revenue and idle capacity.

The six directions a shift can take

Underneath all the patterns sit six dimensions: who you serve, what you sell, how you get paid, how you reach the customer, who you work with, and how costs behave.

The companies that succeed move two or three of them on purpose. The ones that fail attempt all six. The six are covered in six fundamental shifts.

Run it as a bounded process

To make it more than an afternoon, give it four phases with one result each: understand the current model, explore three to five alternatives, design one of them in detail, and test it on real customers before investing.

Two to four weeks in total, with a decision at every handover rather than a running discussion. The phases and a worked example are in Rethink NOW.

Test the smallest thing first

The most expensive mistake is building the whole new model and only then finding out whether customers want it.

A service company with twenty employees wanted to move from project sales to retainers. Instead of building it, they called ten existing customers and offered a six-month agreement priced on what those customers already bought. Seven said yes.

Two weeks, no investment, and an answer worth building on. Had they built first, they would have spent six months learning the same thing.

Get started this afternoon

  1. Draw the nine boxes as they actually are.
  2. Ask the four risk questions and agree which box hurts.
  3. Pick two patterns that address that box, and draw them as complete models.
  4. Write down the smallest test that would settle it, and who you call on Monday.

Point four is the entire difference between a model that gets changed and one that gets discussed.