I just sent an email to my boss, who was asking for an estimate on when my team could have a certain feature deployed to production. This is an extremely common scenario in the s/w dev biz. Instead of going to my team to ask them for an estimate, here& #39;s what I did. (thread)

In the past months, I& #39;ve been collecting data to measure my team& #39;s change lead time to production, which is recognised as one of the four key DevOps metrics according to @Stelligent. Using this data, in Excel I created a histogram with a pareto line. From it I was able to examine

that e.g. roughly 90% of changes were deployed to production within seven weeks, and 99% of changes were deployed within 14 weeks.

So that& #39;s what I wrote in my email to my boss; that based on historic data, there& #39;s a 90% chance that we& #39;ll have it ready in prod within seven weeks

No reply as of now, but I& #39;m curious to read his response to it.

Anyway, this approach I wouldn& #39;t even have considered, if it weren& #39;t for inspiring people like @WoodyZuill and @duarte_vasco. Thank you both for helping me see the world differently! #noestimates