Estimation in software development is challenging, but there’s no other way. Or is there?
In this explainer, we look at methods that use past data – cycle time and cycle time distribution – to predict future outcome. These methods can be tested in parallel with your current way of estimating, until you feel confident that they work – or you find out they don’t work in your case. Their advantage is in less time spent estimating and negotiating estimates, and open the path for continuous improvement based on data. As we’ll see, they also have disadvantages, so definitely watch this and let us know what you think and if they would work for you!
Links:
- Vasco Duarte on fallacies about estimates https://twitter.com/duarte_vasco/status/1655497557869572099?t=JL-RGW1mYx6ZGMVk_t5TKA&s=09