Apologies I only got around to reading your article properly just now. Thank you for sharing, it was a great read.

However, we seem to disagree. Especially on your quote “Time is the wrong measure of success”. In my opinion, time is precisely one of the things that determine success. Time is, besides money, a cost and therefore must be minimised. A well-engineered Blackberry app that took 5 years to build and in that time became obsolete is clearly not a “success”. On the other hand, a buggy website that was made within a month is not a success either. Neither is a mobile app that solves no one’s problem. Success is a combination of providing the most value for the lowest cost.

Just like in any other industry there are constraints. As engineers, we don’t have the luxury to work on a piece of software forever. In most cases, it needs to get out into the wild as soon as it’s minimally viable. Otherwise, the cost is going to increase for whoever is paying us.

Also, you’re proposing that we should not be estimating at all. But that is clearly not going to work in some fields that have constraints on time. For example, farming software that needs to be done before the next harvest season, a game that needs to be done before Christmas and a new fashion website that will lift from the current hottest trend. In these cases, we need to know if the current project is on track to reach the deadline. Because if not, additional measures need to be taken.

So if we’re not going to use relative estimation, how are we supposed to answer the above question?