Estimation: converging to useful estimates

Estimation is a fascinating area in science and engineering, and it is increasingly coming to the fore in other fields. With healthcare projects becoming increasingly complex and driven by growing regulation, estimation is a mix of art and science that needs to be mastered.

Accurate estimates are within your reach

When one doesn’t know, one needs to estimate, which means arriving at an educated guess. This applies to costs, timelines, resources, demand for a product or service, size of a facility, and so on.  In this post and others to follow, the emphasis is on estimation as it relates to project management.

Quite often, an estimate is based on judgment or “gut feeling.”  Perhaps counter-intuitively, this type of guess can be off-base even in the case of experts and thus, as a norm, should not be relied upon without some form of verification.  A more reliable and bias-free guess can be arrived at by collecting and analyzing historical data and building mathematical models of varying sophistication to validate one’s assumptions. Eventually, a properly tested model can be used for planning purposes.

Arrival at an accurate estimate is an iterative process, in which what is known as the “cone of uncertainty” — think of a cone laid on its side with the base to the left — as to the range of the estimate is progressively reduced over time by collecting more and better data.  This means that, while an estimate can be expected to be off early on in a project, it should become much less so as time goes on.  The implication is that estimates should be refined constantly with a relentless focus on project data, and not just done on a one-off basis at the beginning.

Too often an estimate is done at the beginning of a project and is never reworked. Furthermore, because this initial estimate is quite likely and justifiably widely off the mark, when a project post-mortem is carried out, the wrong lesson is being learned by many that estimation is not worthwhile. It is, if done properly.

 

Share

Leave a Reply

Your email address will not be published. Required fields are marked *

* Copy This Password *

* Type Or Paste Password Here *