Gartner’s Magic Quadrant is the most influential non-financial business research document. In the late 1980s, it was a quick and dirty stalking horse to provoke discussions. Today it is an extensive and yet highly limited process, based on the quantification of opinions which are highly qualitative. The early evolution of the MQ tells us a lot about the challenge of industry analysis broadly, as well as the specific process by which the MQ was born.
Later this week I’ll be speaking about the origins of the MQ to the Society for Social Studies of Science and EAAST conference in Barcelona. In the talk, I’ll be stepping through four main ideas. Starting from Grant Blank’s useful theoretical study of product reviews, I’ll argue that the MQ changed fundamentally in its first few years. Grant finds that reviews come in two forms:
- Connoisseurial reviews give the qualitative judgement of an expert reviewer
- Procedural reviews give the outcome of a comparative testing process.
The Magic Quadrant has certainly moved some distance from the first of these forms towards the second. Although Gartner only started to use the term “Magic Quadrant” in public from 1994, similar charts appeared in its 1980s research as the firm gradually adopted desktop publishing software. However, such charts were clearly qualitative. The example above, for example, showed Gartner’s anticipation of how one market would evolve in the future. This is risky analysis: the future is not known and its quantification is very hard to do at the relatively granular level of quickly-changing vendors in a highly discontinuous market.
These early diagrams would have been impossible without Gideon Gartner’s commitment to stalking horses. He described them as tables or diagrams that allow imprecise ideas to be shared so that they can be later refined and enriched through wider discussion and criticism.
These diagrams seized readers’ attention. To many Gartner clients, who found the written research too extensive to master, the ability of these diagrams to summarise a large amount of insight more than compensated for their qualitative and somewhat contingent foundations.
This demand was a significant opportunity for Gartner in the early 1990s, which was struggling with growth. Saatchi and Saatchi, which had bought the firm for $90m in 1988, sold it for $63.4m two years later. One of the challenges for Gartner was to deliver a consistent research experience with a qualitative methodology which depended on highly uneven expert analysts. In 1991, Manny Fernandez left quantitative research firm Dataquest to become Gartner’s CEO. Under Fernandez, MQ-type documents became much more common. Professionals working at Gartner at that time have told me of a period of astonishing growth in which the firm has to master the rapid onboarding of new analysts and salespeople, in particular outside the USA. My conversations have not been exhaustive, but they outline the following scenario.
The wider adoption of the MQ by Gartner services rapidly shifted its time-orientation. MQ’s became ways to show markets as they were rather than showing future direction. Staff working at Gartner at the time tell me that the MQ allowed new analysts to be brought on board more quickly, and to produce compelling research outputs even before they fully mastered their market knowledge. Over the following decades, the MQ has gone through many methodological shifts, most aggressively in 2005. Now the MQ is much more like Blank’s procedural reviews, with extensive data-gathering placing substantial workflow burdens on vendors and producing analysts who could move more easily from segment to segment.
For more insight into the MQ, our Development Workshops allow AR teams the ability to change the way they respond to MQs and produce better outcomes.