Write the case study before you have the case

Before you build the next thing, write the case study you hope to publish about it in six months. Invent the customer. Invent the numbers. Write it properly, as though it were true.

This takes about an hour and it will tell you more than a month of planning.

Why it works

A case study has a rigid shape, and the shape is where the lying gets exposed:

• Who they were, and what state they were in.
• What they tried first, and why it did not hold.
• What changed, specifically.
• What that was worth, in a number they would recognise.

If you cannot fill in the fourth part without waving your hands, the thing you are about to build may not be worth building. That is a useful hour.

Write down the number

Most invented case studies fall apart at "what that was worth". You find yourself writing "saved significant time" because what actually happened was "saved about ten minutes, once".

Make yourself write a figure. Four hours a week. Thirty more people in the room. Two hundred pounds of catering not thrown away. Then ask whether anybody would change what they do for that figure. Often the answer is no, and you have just saved yourself a quarter.

It also gives you the brief

Once the fictional case study reads well, you have your specification. You know the before state you need to detect, the change you need to produce and the number you need to be able to show. Build towards the story and you tend to build the thing that produces it.

Build first and write afterwards and you end up hunting for a story that fits what you happened to make.

Then go and find the real one

You are not going to publish the invented one. You wrote it so that you know what you are watching for, and so that when a real customer does that exact thing you notice, and you ask them for the numbers while it is still fresh.

Most companies miss their best case studies because nobody was watching for them. You will not, because you wrote it down six months early and have been waiting for it.

One rule

Never publish the invented one, and never show it to anyone outside the team without the word "hypothetical" on it in large letters. It is a thinking tool. The moment it leaves the building as though it were real, it is something else entirely.