There are many different ways to estimate software - depending on the timeframe and quality constraints. I’ve refined a number of techniques - and often find the best way is to combine them to fit the situation.
In this example, I have applied a simple risk model on top of a very lean function-point analysis. The accuracy was important - so tee-shirt sizing was not appropriate. In the end, the estimate was within 10% of actual at go-live. So, good result!
©2019 Jonathan Visser
Template by Bootstrapious.com
& ported to Hugo by Kishan B