You didn't lack discipline. The course was the wrong length.

Most people who begin an online course never finish it. Depending on the study and the platform, completion is routinely reported somewhere between the low single digits and the low teens per cent — and for open, free, self-paced courses it sits nearer the bottom of that range than the top.

That number is usually presented as a fact about learners. Motivation, attention spans, the discipline required for self-directed study. It is occasionally presented as a fact about the internet.

It is mostly a fact about course design, and specifically about length.

Length is set by the format, not by the subject

Ask what determines how long a course is and the honest answer is almost never “how long the material takes to teach”.

A course is priced, marketed and compared on duration. Listings advertise hours of content the way a product advertises capacity — forty-two hours, one hundred and eighty lessons — because length is the most legible proxy for value that a buyer can compare before purchase. A shorter course looks like less for your money even when it is more.

That creates a direct incentive to pad, and padding is easy: add foundations, add recap, add a slower ramp, add the module on the adjacent tool. None of it is dishonest. All of it is defensible in isolation. And the aggregate effect is that the course is long for reasons that have nothing to do with you.

Then you fail to finish forty hours of material, roughly twelve of which you needed, and conclude something about your character.

Abandonment is usually rational

Here is a more useful way to read the moment you stop.

At any point in a course you are making an implicit trade: the remaining cost — hours, attention, evenings — against the expected remaining value. You keep going while the second exceeds the first.

Now consider what a poor fit does to that ratio. Every chapter that covers something you already know adds cost and returns nothing. Worse, it degrades your estimate of the future: three consecutive chapters you didn’t need is strong evidence that chapter twelve will be the same, and you are correct to update on it. The expected value of continuing falls, and at some point it crosses the cost.

You did not run out of willpower. You performed a reasonable calculation with the information available and got a defensible answer. The information was just about a course that was fitted to somebody else.

The restart loop

There is a specific failure that eats more time than outright abandonment, and almost nobody names it.

You leave a course at chapter five. Three weeks later you come back with good intentions. You cannot remember exactly what chapter five covered, you cannot tell which of the earlier material has faded, and you have no way to check cheaply — so you do the safe thing and start again from the beginning.

Do that three times and you have watched chapter one three times, chapter five twice, and chapter twelve never. The subjective experience is of having put in real hours on this subject, repeatedly, and still not knowing it. Which is accurate. The hours went into the part you already had.

The loop is caused by a missing measurement. You have no representation of what you know, so you cannot resume — you can only restart. Progress bars do not help here, because a progress bar records which pages were open, not which ideas landed.

What actually helps

Some of this you can do today, without anything changing:

• Decide the exit condition before you start. Not “learn Kubernetes” but “be able to debug why our pod is not starting without asking anyone.” A goal with a testable end lets you stop on purpose, which is completely different from stopping by attrition.
• Treat the syllabus as a menu. You are permitted to take four modules of twelve. The instinct that a course must be completed in order is inherited from school, where it was enforced by assessment. Nobody is assessing you.
• Accept the holes. Skipping creates gaps, some of which will bite. Being bitten by a specific gap and going back to fill it is a fast, targeted, highly memorable way to learn. It is strictly better than pre-emptively studying twenty things to avoid one of them.
• Read the last module first. If the destination is somewhere you already are, no amount of discipline will make the middle worth it.
• Judge length as a cost, never as value. “Forty-two hours of content” is a price. The advertising presents it as a quantity of goods received; it is a quantity of your evenings required.

The part you cannot fix from your side

All of that is coping strategy, and it should be, because the underlying problem is not one a reader can solve.

The material is fixed before you arrive. It was sized for a market rather than for a person, padded by an incentive that has nothing to do with teaching, and it has no way to know that its first three chapters are, for you specifically, pure cost. It cannot skip them, because it does not know you already have them and has no mechanism to find out.

Given all that, a nine per cent completion rate starts to look less like a verdict on human discipline and more like an entirely predictable property of selling one fixed-length artefact to everybody.