We Need More Precision, Says #NoEstimates

It is believed that the Sami languages of Norway, Sweden, and Finland have as many as 300 words to describe snow and ice. Anthropologists have long argued that if you live in the Arctic Circle, you probably need more than the dozen or so words for snow and ice than we need further south. I once watched a documentary that followed a mountain rescue team. One mountaineer described in great detail the difference between different kinds of snow and how some forms are more prone to avalanches – read the first couple of paragraphs from the Wikipedia page on this subject reveals a colourful collection of descriptive terms that have been introduced to encapsulate concepts from the technical domain of avalanche study.

So if you work in technical domain of software development, you probably need a vocabulary that expands beyond the word “estimate”.

The word “estimate” is a little like the words “snow” or “ice”. It captures quite well a very large category of things, but it cannot be used with precision. “Snow” refers to individual snowflakes as well as the collective of all snowflakes laid on the ground. Estimates, too, covers an incredibly broad definition from very lightweight relative techniques, to clever calculations, to projects, and lots more. In common use, it is also mistaken with commitments.

When we talk about estimates at work, we must ensure that our language is precise. We can’t supply an estimate in such vague terms that it can be taken for a commitment. When we make a commitment, it should be clear that we are ready to do whatever it takes to meet that commitment – so having an estimate construed as a commitment damages our profession.

Equally, if we are talking about a specific technique employed in estimation, we should ensure that our methods are transparent. If we have used a projection, that should be well known. If we are at such an early stage in the process that all we can do is apply our instinct and prior experience, that should also be known. The provision of numbers without this clarity is also damaging to the reputation of our profession. Imagine misleading someone into thinking a gut feel number was a confident projection, whether by accident or not the result is the same.

So this is the key revelation for me in the #NoEstimates debate – that we should avoid using vague layman’s terms when precise and unambiguous alternatives are available.