Best-case estimates are still useful

Planning is guessing, as Basecamp says. When we estimate how long a project will take, we don’t usually have an idea of the scope of work. Or the quality at which it needs to be done. Or the manpower available to execute the project — new people might join the team, or existing ones might leave. Or what other tasks they’ll be diverted into, which are usually labeled “urgent” or “critical”. Or what feedback you will get from the real world when you actually deliver your work. As the saying goes, no plan survives first contact with the enemy.

For all these reasons, estimates are guesses. It can take multiple times the estimated amount. Not 30% more.

For this reason, for a while, I refused to do any estimates. I was afraid they’ll be interpreted as commitments and I’ll be held to them.

But now I’ve started to do best-case estimates. I mention in the spreadsheet in big red font, “The project can take multiple times the estimated time” to set expectations clearly. That’s always the case in project management whether they realise it or not. You just want to call it out explicitly.

This does not satisfy stakeholders who want predictability.

Do best-case estimates have any value, then? Indeed they do:

First, you’ll get a sense of how big the project is. For example, I recently estimated a project at 400 hours. Now I know it will take months, not weeks. Such an order of magnitude estimation is still immensely useful. You won’t feel stressed if you know what to expect. You can set expectations with stakeholders that “this will take months”. Then they won’t bug you every week to ask why it’s not done. If they do, you can tell them “I told you it will take months”.

Second, we might decide that we need to hire someone for this. For example, if the company doesn’t have a DevOps engineer, this might be as good a time as any to hire one.

Third, we might even decide not to take up a project we don’t have the time or money for.

Fourth, in consulting, you can translate the best-case estimate into a budget like “This will cost tens of thousands of dollars” so that management can arrange the budget.

Best-case estimates are still better than no estimates. Do them.