Ideas Are Not in Your Head. Go and Find Where They Live.

Four places. That is the answer, and none of them is your memory.

Course ideas do not arrive by sitting down to think of one. That method produces a list of topics — broad, plausible, unrankable — and then you stare at it, cannot tell which is real, and quietly stop. It fails for a structural reason: you are asking yourself to recall problems, and the problems you can recall on demand are the famous ones everybody in your profession already knows about.

The ideas worth having are specific and slightly embarrassing, and they are recorded in places you can actually go and look at.

One: the first six months of your own career

The gap between what your training covered and what the job actually required.

Everybody in a profession has this gap and everybody assumes theirs was a personal failing. It was not — it is structural, because education is designed around a syllabus and work is not. Go back to what you did not know in that first half year. Not the technical fundamentals; the operational reality. How to handle the thing nobody warned you about. What to do when the standard answer did not apply. Who to ask.

That gap is still there. It is being fallen into right now by people six months into the job, and almost nothing exists to help them, because the people who could write it have moved on and forgotten it hurt.

Two: where your profession complains in public

Groups, forums, association boards, comment threads, the review sections underneath existing courses in your field.

This is the highest-yield place on the list and hardly anybody uses it. Read what people actually say. Not what they say they want — what they are annoyed about, confused by, or apologizing for not understanding.

Pay particular attention to complaints about existing education. Somebody took a course and it was a waste of a weekend. Somebody says the exam prep everyone recommends is fine on theory and useless on the practical section. Those are two things at once: proof there is a budget, and a specification for the product that should exist.

If you want to be thorough, buy a competitor's course and go through it properly, noting every place it stops short. That is an afternoon and a few hundred dollars and it is the best market research available to you.

Three: what you paid somebody else to do

Reverse the question. What have you outsourced, avoided, or fumbled through in your own professional life because nobody ever taught you?

Whatever that is, everyone else in your profession has the same hole, and if you have since filled it — properly, through doing it — you are holding something people will pay for. The strongest version of this is the thing that made you feel stupid at the time.

Four: the decision you have never seen written down

Every profession has a handful of forks that get decided badly because there is no honest account of them anywhere. Whether to specialize. Whether to go independent. Whether to take the role that pays more and closes a door. Which of two ways of doing the work to commit to.

You have been through at least one and you have opinions with reasons attached. Nobody has written those reasons down, because the people qualified to write them are busy and the people with time to write are not qualified.

Then the part that turns candidates into an idea

Go to those four places and come back with a list. Now do the work that actually decides it.

Write five problems. Not topics — problems, phrased as something that damages a colleague's working life. If you cannot phrase it as a problem, it is a subject and it goes.

Against each, write what you specifically know that addresses it. Not what could be researched. What you have done. If that column is thin, the problem may well be real and it is still not yours to solve.

Rank them by prevalence. How many people in your profession have this? Prevalence is the single biggest determinant of whether anyone buys, and it is the question people skip because they have fallen in love with problem number four.

Then check you could talk for three or four hours about it, in a structured way, without padding. If the whole answer is twenty minutes, it is an article. That is not a failure — publish it as one and pick something else for the course.

Then kill four.

That last step is the one people will not do. They keep all five, build something that gestures at all of them, and produce a course with no centre that nobody can describe to a colleague.

One problem. One kind of person. The other four go in a file and some of them become your second course, which will be much easier to build than the first because by then you will have an audience telling you which of the four they actually want.