Clearly one of the biggest challenges when starting out an analyst firm is to get the core taxonomy in synch with the market now and in the future. Firstly why build a taxonomy? A key reason is that at key times, I will produce “forecasts” for key markets where I believe I have significant and unique insight. Additionally, from a structural and focus point of view, the taxonomy is important (in my humble view) to provide the consistency and discipline to structurally understand the markets and vendors that will be covered.
After contributing for many years to the IDC taxonomy, I built a taxonomy from the ground up in 2007. It took time. However, by comparison with the current market situation, that was easy. The evolving technology service delivery market has overall been of great benefit to buyers and sellers of IT related services and solutions. However it means that just like a PC, a taxonomy is potentially to readily dated. I am quite sure that the taxonomy I built in 2007 is full of sub-markets that are decreasing in relevance in the current and projected future. It is a tough call for firms who rely significantly on forecasts to keep the relevance, and “history” of their forecasts whilst still being able to evolve a taxonomy to meet the evolving market.
That brings me to the challenge of building a taxonomy for 2011 and beyond. I have had several stabs at it, and the results have not been satisfactory for me. The market is so fragmented, that it is a tough task to bring it together in a cohesive way that will stand consistent. Another challenge is to avoid large loosely defined markets like Systems Integration, that for some are 30% of a market but do not reflect how IT services/solutions are consumed. On the other hand too many markets and the forecasts get loose and lose accuracy.
Naturally I will however keep working on it until I have a taxonomy that suits the emerging markets and the unique way in which we procure services. I am not a fan of importing taxonomies from Europe or the US, though of course these may work for globally modelled markets.
If you have any suggestions as to how taxonomies can be done better and how forecasts can reflect the market better, I really appreciate any feedback.