Posted on April - 21 - 2011
Yes, but what IS it?
For some, the cocktail question is easy. Lawyers only need to specify their field. Doctors, construction workers, and teachers can all clearly explain exactly what they do in a just a few words, saving the listener from having their eyes glaze over in confusion. However, when you work in software estimation, the answer to the question “Oh, what do you do?” is a little more difficult than most.
Ah, the cocktail question
The difficulty lies in two prime areas. The first is that even now as people are moving more and more of their lives online, software estimation still isn’t used in a lot of jobs. The idea itself is slowly starting to take off, but to put it one way I’ve actually heard a coworker say “If you need [software estimation] then you already know what it is. If you don’t know, then you probably don’t need it.” Granted, my coworker’s view was a bit cynical, software estimation is a growing field – Apple’s new iTunes Genius is a very simple version of software estimation – but he does have more than hint of truth in his statement. Software estimation is used to make decisions that are relatively complicated, and those decisions are usually reserved for larger companies for which risks are to be avoided if at all possible.
How complicated? Well, Microsoft’s book on software estimation refers to it as a “black art” which is a bit of a harsh misnomer that casts software estimation as perhaps something that is evil. I can assure you, software estimation is anything but. The term “art” is kind of an interesting fit for just what software estimation does – as no two products use the same process to get the same results.
And of course, none of these actually do explain what software estimation is, or what it does. To put it simply – software estimation is process that takes all these disparate pieces of data, pours them all into a pot, and then predicts the future. Of course, that’s not what I tell people. No, I tell them it’s that I work with fact-based crystal balls. That always seems to go over better.
Related articles
- Data Center Consolidation: Avoiding Surprises and Returning Business Value (galorath.com)
- Data Driven Estimating Part 1: How Data Feeds The Estimation Modeling Paradigm at Galorath / SEER (galorath.com)
- Software Defects in Fielded Software: Cast Analysis (galorath.com)
- Are Unused Software Licenses Costing You Money? (pcworld.com)
Software Sizing, Estimation, and Risk Management is a practical, hands-on discussion of the software estimation, planning and control process. It addresses critical factors that affect estimates, methods for selecting and applying appropriate measures to projects, proper software sizing, processes to identify and manage risk, and best practices to avoid problems and develop successful project plans.
