Comment by Gergely Polonkai on "How do you deal with uncertainty when developing for something new that you haven't done before and set a timeline for it?"

I agree with Dave, with some extensions.

When I have to estimate on something I never did before, I do a bit of research. Even if I’m sitting face-to-face with the client, I ask for some time to look around the Interwebz.

If anyone did that thing in my language, using my framework, I already have a stable base to build on.
If anyone did that in my language with a different framework, I still can do a fairly good estimation.
If I can find a solution in a different language, that will at least double my estimation, but I still have something to work with.

If I can’t find anything similar, I will give a really long timeframe (and probably a huge load of money). If the client asks “Why?”, I break down the problem to small tasks, so they can understand my reasoning better.

However, if I add “because no-one did this before me ever”, the client usually says “we should do it immediately”.

Search Hashnode