Showing Work Before It Is Finished

What goes wrong

You spend two days building the thing, show it, and learn that the approach was not what was wanted. The two days are gone.

Nobody was withholding the information. They had no way of knowing what you were doing, because you were doing it silently, and they could not correct a direction they had not seen.

What early means

Not a half-finished thing. A statement of the shape, before the building.

Before I start on this: I am planning to do it as a single migration that rewrites the column in place, rather than adding a new one and backfilling. That seems simpler but it is not reversible. Does that match what you had in mind?

Three sentences, sent on the first morning, and they catch almost every wrong direction there is.

For a document, five headings. For a piece of analysis, the question you are going to answer and where the numbers will come from. For a design, one rough screen rather than six finished ones.

Why it does not make you look unsure

This is the fear, and it is backwards.

Stating an approach and the trade-off you noticed in it is what somebody with judgement does. Disappearing for two days and emerging with a finished artefact is what somebody without a plan does, and it is only indistinguishable from confidence until the artefact turns out to be wrong.

The form that carries this is a decision, not a question. "I am planning to do X rather than Y because Z" invites correction without asking for permission.

What it costs

Nothing, which is the finding in the panel above.

The review took about the same time whenever it happened — twenty-five minutes for the migration, twelve for the document. What changed was how much work was already built on top of the thing being reviewed.

Showing early does not spend more of anybody's attention. It spends the same attention at a point where it can still do something.

Who to show it to

Not always your manager.

The first row is the one new people get wrong. The person who will review your work at the end is the right person to show it to at the beginning, and going via your manager adds a day and a translation.

The habit by month two

It stops being an event.

A line in the channel saying what you are about to do, a draft shared at a quarter done, a question about the approach before the work — by month two these are not announcements, they are how you work, and nobody reads them as uncertainty.

That transition is most of what the second month is, and it happens faster if the first attempt is made in week two rather than week eight.

The worry about wasting people's time

The objection to showing early is that it asks for attention twice: a look at the shape, and a look at the finished thing.

In practice it does not work that way. The review of a shape takes four minutes and the review of the finished thing is faster, because the reviewer already knows what they are looking at.

The third row happens more often than people expect: a reviewer who has agreed the approach frequently waves the finished thing through.

So showing early does not spend more of anybody's attention. It spends slightly less, at a moment when it is worth considerably more.

The exception

One case where finishing first is right: when the thing is small enough that showing it twice costs more than doing it wrongly.

A half-day task with an obvious shape does not need its approach agreed. Sending a message about how you intend to spend four hours is an interruption with no return, and doing it habitually is its own kind of overhead.

The threshold is roughly a day. Below it, do the thing and show the result. Above it, say the shape first, every time — and the longer the task, the earlier the saying.