Saturday, January 06, 2007
Taxonomizing Open Source Analysis
Here are some of the quotes along with my perspective:
- end-user clients often cite the main reason for using Gartner is that it provides a one-stop-shop for research covering most IT and business topics.
First, we need to deconstruct the meaning of one-stop shop from an enterprise perspective. It does mean the following:
- From a procurement/contracts perspective, there is only one agreement and one flat fee.
- From an end-users perspective, the terminology used is consistent. For example, take the word open source. Some analyst firms will stay more pure and use the ten principles outlined at Opensource.org while others will use whatever definition is supplied by the vendor. Consistency across analysts is important.
- We desire to have one place to submit requests for briefings.
- We desire to force all of our requests through a single internal coordinator because of the notion of a seat.
- Take it as a call to the Open Source Analyst community: it should organise itself to provide a single-source (research) distribution (channel).
It is my thought that distribution should be made more organized but not necessarily research. In fact, I would probably argue against it. For example the notion of a Quadrant or Wave is useful in terms of me filtering down a list of vendors to invite in for a proof of concept and is primarily targeted at IT executives and non-technical software architects (A pervasively implemented oxymoron in enterprise environments). Other analyst firms provide insight at a detailed level and simply target a different audience. Normalization could alienate one demographic.
- The idea is not new and implies agreeing on a taxonomy, possibly working together on research pieces, vetting in/out members, etc... .
Let me state with passion that analyst taxonomies are evil. Taxonomies as used in the analyst space tend to be how analysts classify vendors which in many cases is not how we customers use technology. It wouldn't be too difficult to find a software vendor (or a hundred) that believe they are inappropriately classified by analysts. So if vendors and customers agree that taxonomies are evil, let's not go there. Let's instead figure out better mechanisms for tagging.
The notion of working together on research pieces seems interesting at some level but I think regardless of whether that happens or not, I would love to at least see analysts reference the work of others. The only analyst I have seen that has done this has been Brenda Michelson of Elemental Links. This is incredibly transparent of here and does wonders for her credibility...
Links to this post: