Original Post Date: September 30, 2015

‘It’s crazy to think that you’ll be good at estimating things that haven’t happened yet, if you can’t even accurately say how long things that have happened took.”  This quote from the blog post “5 Ways Software Developers Can Become Better at Estimation” (http://www.javacodegeeks.com/2014/10/5-ways-software-developers-can-become-better-at-estimation.html)  speaks volumes.  The current mantra – at least in the Aerospace and Defense communities is data driven estimation.  Estimates without solid project history data to back them up are consistently rejected or approved with substantial risk assigned.  Data collection in the world of estimation is no longer a nice to have but rather a necessity.

But as the author in the above mentioned quote so nicely puts it … how can you predict the future without understanding the past?  There is no magic bullet for estimation and every new project is a new project with unknowns, risks and surprises.  However – every new project most likely has some (possibly significant) overlap with projects previously performed.  Sure there may be a new development environment or manufacturing technique, but seriously if you’re organization is planning on bidding for the project, it’s unlikely that you have no experience in the field or market.  So even if the data collected from previous projects is not spot-on to the one in question – it has to provide answers to some of the questions.  If nothing else, the data you collect from historical projects (after you’ve studied a few) offers information on how proficient (or not) your organization is in its ability to deliver products.

Good data collection is essential to good estimation.  Using an estimation tool, such as TruePlanning, can help to make data collection more efficient and successful.  First of all because tools offer a structure to the data collection process.  The input parameters represent guidelines on the important technical parameters that function as cost drivers in a project such as software’s functional size or the experience level of the technical staff.  They also provide a repeatable scale with which to measure such technical parameters.  The output categories offer a structure and repeatable way of collecting output data.  TruePlanning also provides a methodology through which calibration can be used to turn a general purpose model into one that is fine tuned for the specific way an organization attacks their projects.