Gartner: Early Research Process (RP), Part 2 (including MQs)

Gideon Gartner on Magic QuadrantsIn Part 1 of this post I mentioned some of the sources for my research process ideology, resulting in a 1980s ‘research process’ for Gartner which included six major elements: Surveillance, Pattern Recognition, Stalking Horses (including the Magic Quadrant), Search, Document, and  Strategic Planning Assumptions (primarily for conferences); while perhaps not followed  explicitly, these elements  contributed to our research culture:

  1. ‘Surveillance’ was simply a qualitative requirement; we expected our analysts to  be sensitive to all appropriate inputs; we wished our analysts to read extensively, attend relevant meetings, visit with vendors, and pursue meaningful  conversations with  clients, peers  and supervisors,  all with the purpose of optimizing the  preconditions for pursuing relevant research and documentation.
  2. ‘Pattern Recognition’ was  a talent or skill which we hoped our analysts would have when hired (after surviving our tough interviews) but even if not, would be developed.  When added to  ‘surveillance’(see above),  I felt our analysts should attempt to identify ‘patterns’,  leading to more original and provocative research conclusions than our competition. It’s always difficult  to explain how one should precisely recognize patterns, especially with some iota of precision; worse, the concept itself is  abstract, for example the orientation of Wikipedia’s definition is  ‘machine learning’.  I did find  one description  which approached what we were attempting (see note at end of this post), yet I believe that most analysts understood that  keeping their eyes open to the broad IT spaces they might recognize certain ‘patterns’ which  could lead to original interpretations, and of course to our goal: original conclusions!
  3. ‘Stalking Horses’ (sometimes called ‘Straw Men’). This did not become a critical part of our research culture until 1987 (our eighth year).  I don’t recall precisely how  I came up with this one,  but the idea was  to create a graphic or tabular presentation, hopefully including quantification of the idea or concept which would then serve as a basis for further study… and inquiry….and discussion…  perhaps finally resulting in an approved consensus view at our research meetings.  My personal interpretation of the Stalking Horse term was that  when a concept or issue would be elusive (as almost all information technology issues are), we would sneak up with our ‘stalking horse’ (a table or diagram or graphic which we would draw), and close in towards a solution while in turn initiating colloquy with our peers! We could then iterate around our varying opinions, data labels and values, until we hopefully reached reasonable consensus. Our databank would then be enriched with a new-found Gartner conclusion!
  4. After I introduced the concept, one of our outstanding researchers (Bruce Miller, a superior Gartner analyst)  justified the term ‘Stalking Horse’. The story: Miller’s  initiative, researching and  finding an early engraving by a gentleman named ‘De Bry’ of an even earlier ‘Le Moyne’ painting which showed American Indians using animal hides (likely horse-hides) to stalk deer without alarming them, led us to the term ‘Stalking Horse’! More specifically, Bruce’s search found that the words actually appeared in print during the early 1500s, describing  ‘a façade or blind used to mask a true purpose’. So what? Well, just as early hunters used these tools to sneak up on elusive quarry, in our context ‘stalking horses’ would be graphics and techniques to ‘sneak up’ on market growth rates and dozens of other business variables!

    This concept became an important part of our research process, so important that it permeated our entire research culture! I began documenting and promoting  the Stalking Horse concept, and one day imagined that Gartner might benefit if I edicted “the horse” as our company’s mascot!  Perfect timing: we had just completed our new Stamford CT building,  I  immediately hired an art consultant to help me choose a focused  collection of paintings and sculptures of horses; I  purchased a dozen paintings and sculptures  which we placed throughout our spanking new headquarters! Even my daughter Sabrina, an artist as well as  employee, embraced the concept by producing a huge and beautiful pastel for us,  somehow depicting both the horse and our culture (Sabrina was and is a horsewoman)!

  5. ‘Search’ .This next element of our system was simply the proactive research which we expected analysts to pursue,  collecting all available evidence to develop useful analysis for presentation and documentation to clients.
  6. Document is of course self-explanatory; developing our research conclusions would be useless if not disseminated to  significant subsets of our clients, but writing  and meeting our documentation standards, and in timely fashion, was  challenging  to many analysts.
Note re the ‘Magic Quadrant’, or MQ, as one of our ‘Stalking Horse’ methods:  around 1987/88 ,  I was working on further enhancing the Gartner Group Research Notebook,  and happened to introduce the  ‘stalking horse’ graphic as  one technique which our analysts  could occasionally use when presenting at research meetings to support a conclusion.  This initiative was meant to be for internal discussion only,  because the method seemed  an oversimplification (yet  an interesting  starting point for  certain confrontations  which  would be  challenging and educational at our meetings).  I do not recollect ever allowing the publication of an MQ , but during the mid-’90s and well after I left Gartner,  the  MQ  feature grew to be a major  deliverable  for many of its clients!

I acknowledge that I had left money on the table, as  MQs grew in external popularity ( and considerable dislike). In a future post, I expect to discuss the damage which this tool can create,  sometimes leading to  questionable corporate decision-making.


Leave a Reply