Automated counting is fairly new, which has some people skeptical about its “accuracy." Given that there is no universally agreed unit of measure for software size (no block of metal in a French standards institute), the real test of a new way to measure is if it has broadly linear proportionality to things we care about (e.g. software project effort), and if they can be measured consistently. So, over the last few years, DCG and CAST have worked together to perfect the algorithms and processes that we use for the automated counting of applications.
Automated counting creates consistency in the counts because once the algorithms are set up and the applications are calibrated, each count will be carried out the same way.
Of course, DCG provides both manual and automated counting services, but our automated counting services are carried out via our partnership with CAST.
Both manual and automated counting techniques can be useful and consistent for sizing software, but there are times when one technique may be better for a particular situation than the other. For instance, if you have a lot of applications to count repeatedly, automated counting may be a more economical choice.
We have a series of videos up on our YouTube channel that we’ve created with CAST to provide a better understanding of automated function point counting, such as where it can be useful and where it cannot be used (e.g. measuring size before source code is available).
Watch the video below to hear from Mike Harris, President and CEO of DCG, as to when automated counting may be preferable over manual counting and what benefits exist in utilizing automated counting techniques with CAST.