How much Enterprise software sits on a shelf unused?
Unused because it didn’t work as expected or perhaps it required a higher level of skill than perceived. Maybe it was bought by one team and ignored by the others?
In reality it is probably a mix of these with a few other challenges thrown in for good measure.
With open source and community-driven software – will this impact the value chain?
Be under no illusions open source often has a commercial basis so what levels of support or frequency of release might you expect. This will vary dramatically across companies, product and geography and in many cases you will find the contributors are also active in the support communities responding to you before any commercial response.
Then consider the maturity, or perhaps better expressed as enterprise understanding, of the contributors. Do the people designing, writing or reviewing the code or features have a real background in the enterprise or an understanding of context. As someone who has spent two year at Chef I believe this is one of the more significant and, as yet, unresolved challenges. It is UNLIKELY that the community does have exposure or understanding!
Do not under-estimate the expectations of the community. Open source fosters an expert audience – expert but often lacking emotional connection or empathy. The author has observed contributors who no appreciation of the expertise available in a standard IT department and have never been exposed to IT service management tools or practices. This is not surprising but it is a significant consideration when you buy in to their ecosystem. Beware of features or processes that don’t or cannot fit with the way you run your business, examples include:
– inconsistent user management / directory integration capabilities
– ignorance of IT risk and mitigation
– poor documentation
– over-reliance or ignorance of tooling and process such as orchestration
– the ‘you can write that’ syndrome
– solutions that actually cause problems including false positives and ignorance of the need to use exception-based notifications when working at scale