The case for open source experimentation before buying licenses

Competition is growing more fierce and most markets are being disrupted. To get started and deal with uncertainty it makes sense to shift toward a more lean and frugal approach.

Consider the following question: would one buy insurance before the car? Probably not; so why would one buy licenses for support and bundled open source software, if one does not yet know if it is actually needed, or if that software will provide the expected value.

Traditionally, in an enterprise environment, it used to be the case that value could not be unlocked without licensing expensive software, and it was reasonably straightforward to determine what was needed; alternatively systems evolved extensively over time. With the advent of big data, machine learning and the explosion of related tools, all this has changed. The landscape and development road map are now more varied, more unpredictable and changing faster than ever.

Now, because of the proliferation of open source software and strong communities behind big data and artificial intelligence, everyone has the ability to download and install everything that is needed. Everyone can freely run the software commercially without paying licenses fees or support. This allows for the building of prototypes for the highest value use-case(s), while at the same time vastly reducing the risk to an organisation’s bottom-line. Build a minimal viable product that provides value. Build something that will either improve revenue or reduce costs. Only, once the value has been proven, it will be clear what is really needed to grow, operate and manage the system.

Once a point is reached where a need for licensed enterprise support or features is real, then one can buy it knowing that there is significantly lower risk of making a costly mistake. Taking on costly capital expenditure, without sight of the potential ROI, becomes an unnecessary risk with open source experimentation. It simply does not make any sense to incur an expense until one knows what is really needed and what actually works.