How can I tell if what I am describing is a problem or a solution?

The test for telling a problem from a solution: a problem can be photographed. It exists in the world, in a place, at a time, with a person in it. A solution cannot; it exists only in your head until you build it. The same technology aimed at different problems becomes different products, which is why the distinction is the whole game.

Why does the same technology become different products for different problems?

The test for telling a problem from a solution: a problem can be photographed. It exists in the world, in a place, at a time, with a person in it. A solution cannot; it exists only in your head until you build it. The same technology aimed at different problems becomes different products, which is why the distinction is the whole game.

What does it mean that the problem should be visible before the product exists?

The test for telling a problem from a solution: a problem can be photographed. It exists in the world, in a place, at a time, with a person in it. A solution cannot; it exists only in your head until you build it. The same technology aimed at different problems becomes different products, which is why the distinction is the whole game.

You can photograph a problem · Product Discovery

§Here is the whole of Part I in one test. Could you take a photograph of it?

§A woman at a restaurant table, phone in hand, looking for the waiter who has not noticed she wants to pay. A developer with two windows open, copying a block of code from a chat into an editor and back. A founder at eleven at night, scrolling a spreadsheet to find the deal they forgot to follow up on. Each of those can be photographed. Each exists in the world with a person in it, at a place, at a time. Each is a problem.

§Now try: "a loyalty program." "An AI pair programmer." "A CRM with reminders." You cannot photograph any of them, because none of them exists until someone builds it. They live in your head. They are solutions.

§ 3.1Why the test matters#

§It matters because the two things get confused in language constantly, and the confusion is expensive in one direction only. Mistaking a solution for a problem sends a team to build something and call the building research. Mistaking a problem for a solution is harmless; it just means you have more to learn.

§The confusion is built into how people talk. Ask a customer what they need and they will name a , because products are the nouns they have. "I need a better calendar." That is a solution wearing a problem's clothes. The photograph behind it might be a person double-booked in front of a client, or a person who cannot see their partner's schedule, or a person who blocks time and then ignores it. Three different problems, three different products, one sentence.

§Teresa Torres draws the same line with different words: an opportunity is a need, pain or desire that exists in the customer's life; a solution is what you might do about it. The tree she builds hangs solutions under opportunities for exactly this reason, so that a team can never look at a solution without seeing the problem it claims to serve.

§ 3.2Same technology, different problems, different products#

§The clearest demonstration of the distinction is what happens when one technology meets several problems.

§A large language model is a technology. Aimed at the developer copying code between windows, it becomes an editor that writes alongside you. Aimed at the person who does not trust search results, it becomes an answer engine with citations. Aimed at a customer-service queue, it becomes an agent that resolves the ticket. Same capability, three companies, three products with almost nothing in common, because the problems had nothing in common. The technology was never the product. The photograph was.

§ 3.3What the photograph gives you#

§Three things. It gives you a place to go: the problem is happening somewhere, and you can be there when it does. It gives you a way to know when you have understood: you can describe the picture in enough detail that someone who was not there recognizes it. And it gives you a test for every solution that comes later: does this change the picture, or does it just sit next to it?

§Part I is done. Part II turns to the people who built the methods most founders inherit, and asks of each the same three things: what it gave us, what it assumes, and where it breaks at one to five people.

3. You can photograph a problem