Some people say the Gartner’s Magic Quadrant (MQ) is the most influential diagram in the business world. Each quadrant positions the major technology solution providers in a stated market according to their vision and execution. Many vendors dedicate huge resources to improving their position and, while low scores make some vendors despair, even a modest position on a Quadrant adds a vendors into consideration by a whole layer of prospective customers.
I recently spoke about it with Michael L. Schneider Ph.D., who spent seven years as Vice President at Gartner Group, where he established Gartner’s Advanced Technology practice. Mike has an interesting observation: The absence of numeric scores for weights, individual components and the two figures of merit (ability to execute and completeness of vision) indicate that the positioning of companies is more qualitative than quantitative.
Most solution providers take a pretty opposite approach they focus only on the variables that are explicitly states in the MQ, forgetting that it’s a piece of “magic” in which the analysts’ qualitative judgement, rather than an Excel macro, produces the final outcome. While Gartner provides criteria and general weights against which they evaluate different vendors, these factors may be of a subset of all the criteria used by a Gartner analyst in determining absolute or relative positions.
“When the the reported strengths and cautions included in reports does not directly coincide with the criteria specified as factors contributing to the positioning of a vendor” says Dr Schneider “it further indicates other, unidentified factors contributed to a vendor’s position. In fact, the analyst may be unaware that these factors may contribute to a vendor’s position.”
Since Gartner can’t publish the specific data and weighting used in any particular quadrant, the question arises: is it possible to determine underlying factors that can improve a vendor’s position?
Dr Schneider’s fascinating suggestion isn’t one that many marketing manager will have thought of: “The answer can be derived by examining a combination of stock market technical analysis, the sophisticated models used by internet vendors in targeting potential customers and neural nets used in artificial intelligence. These techniques rely on behavior, not fundamental data. For example, internet vendors have models of user behavior so accurate that they can derive a good estimate of a woman’s delivery date. In one case of which I am aware, these models determined a woman was pregnant before she was aware of her condition. The data used in these predictions may not even appear to be related to the issue examined.”
I think this is a very interesting, paradigm shifting, approach which might fit an evidence-based analyst relations team. With a method like this, it is possible to produce a set of criteria for a MQ that reflects the analyst’s conscious and unconscious criteria and decision making process. Having determined an underlying model, it is possible to perform “what if” analyses of changes made to reposition a vendor in the MQ. So, for example, information provided to Garrtner should focus those areas where the model indicates the greatest benefit is derived. This may include non-fundamental information, such as the way the information is presented, where or when it is presented and the individual who presents it and the relationship between the analyst and the presenter.
Drawing on his years at Gartner, Schneider explains that “when an analyst first starts to create a MQ, the first thing that occurs is the conscious and unconscious the creation of a model company. In psychological parlance: a cognitive map. As an aside, this strongly contributes to successful and unsuccessful user interfaces. Once this map is created, it is used to judge different vendors. Not only does it incorporate the analyst’s background, but can be strongly influenced by vendor provided information. This, in turn, influences the content and method of the presentation of a company and it products.”
The object of analyst relations is to influence the analysts, but influencing the model an analyst uses to measure vendors and products must, surely, have an even greater influence. Schenider’s approach opens up the possibility of influencing the positioning of vendors in the analysis space as well as including a vendor in a specific MQ.
This article first appeared on the Loudhouse blog.