The real cost of building one interactive session

Part of a single argument: the engagement gap is a workload problem.

Ask someone how long it takes to prepare an interactive session and they will estimate the part they remember. The number that comes back is usually the writing: an hour, maybe two.

The actual figure is normally two to three times that, and the gap is not padding. It is real work that genuinely happened and did not get counted, because it did not feel like the task. This is why engagement plans slip so reliably, and why the person who slipped feels vaguely as though they wasted the week.

The visible hour

Writing the questions and building the activity. This is the part people picture when they think about preparation, and for a topic they know well it is genuinely about an hour.

If this were the whole cost, engagement would not be a problem anywhere. An hour is findable. Nobody is dropping interactive elements because they cannot find an hour.

The invisible three

Finding the source material and checking it is still current. Rebuilding something a colleague already made, because you could not find it or could not tell whether it was the current version. Testing that the activity works on a phone rather than only on your laptop. Fixing the version where the figures were out of date. Sorting out access for the two people who could not get in last time and will otherwise not get in this time either.

None of these are the creative work. All of them are load-bearing, in the sense that skipping them produces a visible failure in front of an audience. Together they are most of the cost.

They are also invisible in a specific way: they do not feel like preparation while you are doing them. They feel like obstacles to preparation, which is why they never make it into anyone's estimate and why the estimate is always wrong in the same direction.

Which half is genuinely necessary

Writing the questions is irreducible. It is judgement about your material and your audience, and it is the thing only the person who knows both can do.

Nearly everything else is search, duplication, reformatting and logistics. Those are categories software is good at, and they are the categories consuming the majority of the budget.

The distinction matters because the usual efficiency advice targets the wrong half. Tips for writing questions faster save minutes off an hour. Not rebuilding what already exists saves hours off three. If you only get to fix one thing, fix the one where the money is.

Why the estimate is always low

There is a specific mechanism here worth naming, because it recurs.

People estimate by recalling the last time they did the task, and memory preferentially retains the parts that required thought. The forty minutes spent hunting for a document does not encode as work. It encodes as friction, and friction is forgotten.

So the estimate is not a lie or an optimism bias in the usual sense. It is an accurate recall of an incomplete record. The fix is not to tell people to estimate better, which does not work, but to measure once so that everyone has a real number to reason from.

What to do with the number

Once you know the real cost, two things become possible that were not before.

You can make honest commitments. A team that knows an interactive session costs four hours rather than one stops promising four of them a month, and stops feeling as though it failed when it delivers two.

And you can target the reduction properly. If three of the four hours are search and logistics, then a shared library and a simpler joining flow are worth more than any amount of improvement to the authoring experience. Without the measurement, teams reliably optimise the visible hour, because that is the hour they can see.