The same module for everybody
Practically every industrial site carries the same skew in its training offering. The mandatory module runs forty minutes and is built for somebody who does not know the subject yet. It gets imposed on the engineer who has worked with it for twenty years too.
That engineer clicks through it. That is not unwillingness, it is a rational response to material that tells them nothing new. It costs them forty minutes, delivers nothing, and damages something harder to repair than time: the assumption that training in this organisation is worth the effort.
Adaptive material is the answer to exactly that problem. Not to a wish to do something with personalisation.
Three levels of adaptive
The term covers three very different things, and the difference in price is considerable.
A pre-test with exemption. Somebody takes an assessment first. Whoever masters the components skips them and does only the rest. Simple, cheap, and responsible for the overwhelming majority of the gain.
Branching by role or by answer. The material splits at fixed points. An operator gets a different route than a maintenance engineer, or somebody who gets a question wrong receives an extra explanation first. More laborious, because every branch has to be made and maintained.
Genuinely adaptive. The system tracks how confident somebody is per subject and picks the next question or explanation to suit. This is what suppliers mean when they use the word in a demonstration, and it is the form that demands the most content.
The content you need for it
This is where most implementations stumble, and it is a calculation you can do in advance.
Adaptive material can only shuffle what exists. If you want three difficulty levels per subject and enough questions to stop people seeing the same one again, the question bank quickly becomes a multiple of what you would build for an ordinary module. With a branching route the same applies to the explanations themselves.
That does not make the investment unwise, but it does move it. You are not buying a clever system, you are buying content a system can do something with. An adaptive engine with too little material underneath behaves like an ordinary module with extra clicks.
Count the maintenance in as well. When the procedure changes, it changes in every variant.
When it pays off
Four situations where the sums almost always work out.
Large groups with widely varying experience, where the time of experienced people is expensive. Mandatory repetition of material most of the group already masters. Onboarding in which career changers with relevant knowledge arrive alongside people with no background. And certification programmes where demonstrable mastery per component counts.
Outside those situations it is often more honest to put the money into better basic material. A short, sharply written module everybody works through beats an adaptive route through weak content.
The cheap version that does nearly the same
For most organisations the wisest step sits here, and it needs no adaptive platform.
Put a pre-test in front of the module. Whoever passes gets an exemption and a note in the record. Whoever fails does the module. Done.
That intervention can be built on an ordinary learning platform, costs a fraction of an adaptive programme, and takes on most of the problem: experienced people stop losing their time, and you get a measurement of what the group genuinely already masters. That measurement is useful in itself, because it tells you which components you still need to offer at all.
If the exemption then turns out to be too blunt, you have a concrete reason to go further. Starting at the expensive form without that reason is doing it backwards.
How this trade-off fits into designing a training programme is in the guide on setting up or renewing a corporate academy. The underlying design principle is explained briefly under instructional design.