When it comes to measuring the effectiveness of a given application development project, value has always been pretty much in the eye of the beholder. In theory, the success of a software project could be determined by the amount of time and money it took to provide a specific function. This has led to the development of a school of thought known as function point analysis that senior IT managers can theoretically use to measure how a given software project is progressing. This then leads to conversations with financial types that are anchored around estimating the cost of a project around the number of function points delivered per dollar invested.
The problem with this approach is that measuring the actual function points in any given application is a tedious affair, so the end result is that nothing gets measured and the financial types get frustrated. Today, CAST Software announced what Jitendra Subramanyam, director of product strategy and research, bills as the first tool for automating the counting of function points within any given application development project.
For developers this tool doesn't add any value in terms of getting the project done any faster. But the people who pay for these projects have been groping for a common nomenclature to measure the rate at which a software project is progressing, and most importantly, how much money is being spent to achieve some level of defined functionality.
As the amount of money being spent on software development and the value of that software comes under more scrutiny than ever, the need to find a way to measure software development projects effectively is only going to increase. For years, most business people were content to treat the development of software as something akin to a mystical process. But with the valuations of more companies than ever tied up in their software assets, it's only a matter of time before the same financial instruments that are used to measure the effective usage of other physical assets are brought to bear on software, as well.