DEV Community

Cover image for If you asked how long a job will take, don't provide an estimate - do this instead.
Alex Pushkarev
Alex Pushkarev

Posted on

If you asked how long a job will take, don't provide an estimate - do this instead.

It is important to understand that even if we use the same word we could mean/understand different things.

When people ask "How long job will take" they may be interested in any of the following:

  1. How long will you mostlikely spend doing the job
  2. What is the best-case scenario?
  3. When the job will be most certainly completed (worst case scenario)?

There're at least 63% chance you misunderstood the question. It might be the case that person who asks you doesn't know what they are interested themselves.

  1. So what you should do instead? Consider these options:
  2. Ask what figure they are interested in (mostlikely, best-case, worst case)
  3. Provide all three estimates
  4. If person needs a single number - suggest them checking 3 point estimations (https://www.youtube.com/watch?v=6U5UpDmI77A&t=115s)

If you want to know more about why estimates are usually wrong and what one can do about them - feel free to check this video:

Top comments (2)

Collapse
 
jmfayard profile image
Jean-Michel πŸ•΅πŸ»β€β™‚οΈ Fayard • Edited

Real issue here!

What I usually do is that I say something like that.

"I don't have enough information to provide a reliable guesstimate. What is this feature worth to you?


Worth spending like height hours


if you agree with it, I can spend two hours (time boxing) investigating the topics in detail. Best case scenario I might be done in two hours if we are lucky. Worst case scenario we spent two hours knowing it's not worth doing.

Collapse
 
dadyasasha profile image
Alex Pushkarev

interesting approach, thanks for sharing!