Ten years ago, many of us thought that the days of companies writing their own software would soon be over. Commercial packages such as SAP and Oracle would become more and more comprehensive and flexible, eliminating the need for much custom code.
But like so many trends in business, the build/buy decision is a pendulum, and Erik Keller at AMR says that, for many companies, the pendulum is swinging back toward the middle. Keller points out three factors that are working in favor of the build option:
- Open source, which gives companies a cost-effective platform and a cheap starting point for their own application development efforts. Examples include open source content management packages such as Zope, Red Hat, and OpenCms, and ERP/CRM packages such as Compiere, Ohioedge, and Anteil.
- Offshore development, which lowers development costs, at least for program coding. I’ve pointed out in the past that when considering offshore development, companies need to consider the hidden costs and the risks of offshoring as well as the labor savings.
- Web services, which allow custom applications to be more easily integrated with existing systems and commercial packages. Although web services have been touted for several years as the wave of the future, the promise is only beginning to be realized.
I agree with Keller’s general direction. However, I believe that most companies, especially small and mid-size businesses, would be best served by exhausting the search for commercial software before taking the plunge into custom development.
This is especially true with major enterprise level functionality. Custom development has its own set of issues and risks, and unless a firm has a competent internal development organization, or has the in-house expertise to manage a third party developer, it’s better not to start down that path.