Opinions on Gartner's infamous Magic Quadrant abound (89k hits on this query) on the internet, but it remains one of the most sought research piece.
So well deserved kudos to Louis Columbus (a former AMR analyst) from CRM Buyer for laying simply the issue in his piece "Gartner's Magic Quadrant May Need New Pixie Dust":
- He details the methodology and its flaw
- He advises technology buyers to do their research and not rely solely on the quadrant
Read it, it's well worth it.
Just a few comments:
- Gartner's methodology is (for the best) inconsistent: some quadrants use a dartboard approach (similar to the "beer quadrant" Louis refers to), others use detailed analysis. The server quadrant for instance is backed by the Application Server Evaluation Model (ASEM), which is customisable to match users needs and requirements. They are still not where the METAspectrum was in terms of methodology, but one never knows....
- The publication frequency is also inconsistent, and depending on where the a quadrant comes in a product cycle, a vendor can come out quite well or not. For instance, this one came just before product launches by Hitachi and IBM. The quadrant could have been very different for EMC if it was published in December (the next one is still not published).
- Marketers are lazy. The quadrant is easy. It's not going to go away!