Analytics as an Ideology

If pursuing rigorous, complete, precise, and complex solutions generates unintelligible and unused results, it would seem analytics will be a fad, abandoned because of its inability to deliver on its promise.

Analytics is presented as an objective field of applied statistical techniques. One could reasonably think of analytics as a science. Such a view is not wrong; it is, however, incomplete. Besides these elements, analytics is an ideology. Viewing analytics as an ideology is certainly uncommon. In fact, analysts might reject such a conceptualization. To understand better the perspective that we could consider analytics an ideology, along with its corresponding consequences, it is useful to examine first what ideology is and how it functions organizationally.

What is ideology? For such a seemingly simple question, an answer is remarkably elusive. Understanding ideology, in general, is complicated by the fact that the meaning of the term itself changes based on time and context. Most directly, ideology is a set of interrelated ideas. More constructively, ideologies combine ideas that provide a basis for interpretation and meaning. At its most abstract, ideologies can be understood as a system of falsehoods that are sustained and propagated to sustain a given power structure. It is this perspective that is beneficially pursued as it relates to analytics. Althusser explored elements of ideology as it relates to the reproduction of capitalism. The context he developed is useful, as analytics is so frequently employed in attempts to optimize commerce that its repeated application generated its own field of study, business analytics. As Althusser explained, “ideology exists in institutions and the practices specific to them;” more specifically, “ideology exists…precisely in the practices of…individuals.” This isn’t necessarily an insight of critical import. In analytics, it is likely unproblematic to point out that analysts within organizations implement their tasks consistent with institutional and professional practices, as both organizations and the professions hold unique procedures and protocols. What often goes unacknowledged is that these practices are inherently ideological. More consequentially still, ideology shapes and constrains action. Expanding on his initial point, Althusser noted that “ideology makes individuals ‘act all by themselves,’ without there being any need to post a policeman behind each and every one of them.” This is where analytics as an ideology becomes increasingly worthy of exploration. Analysts, constrained by analytics as an ideology, enact organizational analysis in ways that are potentially contrary to what they think would be beneficial.

For the sake of exploration, let’s assume that individuals are recruited into becoming subjects under an ideology of analytics in a way consistent with what was presented, that individuals become analysts and enact the ideology of analytics. So why is this important? The ideology of analytics, when internalized sufficiently, is focused on values that are potentially aligned only poorly with organizational needs. We can see most clearly this by examining the analytic value placed on rigor. Analytics as an ideology, in pursuit of rigor, often results in false precision and unnecessary complexity. This entails a misallocation of resources, as these elements require additional time. Further, the complicated solution could be unintelligible to those making the organizational decision. In short, analytics as an ideology, in its pursuit of rigor, could lead to results that take longer than necessary to produce and are of less organizational value because they are unintelligible to those required to understand.

At issue here are the contradictions and ambiguities resident in applied analytics. As one develops the skills, techniques, and insights of the domain, one is exposed to an unspoken ideology associated with its use. This ideological foundation is observable anytime one declares what or how something should be done. Normative statements betray ideology. Traditionally, analytics as an ideology places notions like rigor, completeness, precision, and complexity at the core of its domain. While these elements certainly have value in the analytics, it is worth considering if they are sustainable as the core ideology of the profession. Alternatives certainly exist.

I will present here only two alternative ideological values for analytics for consideration, utility and parsimony. Applied analytics is a business function requiring an allocation of resources. For analysis to be organizationally sustainable, it must generate more value than it consumes. This can only occur if the results are timely, implemented, and beneficial. In short, analytics as an ideology benefits from a focus on organizational utility. Related to the focus on utility is the notion of parsimony. The analytic solution benefits from being unencumbered by unnecessary complexity. Analysis can be complex. We derive benefit from ensuring that the solution is no more complicated than it absolutely must be for accuracy. While it is too reductionist to posit analytics as an ideology as a simple tradeoff between competing values, it might be useful to consider the metaphor of a scale on which one attempts to balance rigor and complexity on one side with utility and parsimony on the other. Each side offers something of value. The effective balancing of these ideological elements is the crux of this concern.

Ideology is likely unavoidable. However, one does not need to accept the dominant ideology without thinking about its content and consequence. Perhaps the dominant ideology is well aligned with organizational needs and those of the profession. Perhaps not. If pursuing rigorous, complete, precise, and complex solutions generates unintelligible and unused results, it would seem analytics will be a fad, abandoned because of its inability to deliver on its promise. Making the ideology of analytics explicit is beneficial, as it highlights points around which discussions and improvements can occur. We focus Heath Analytics on developing an ideology that is aligned with the production of results that are organizationally useful and individually engaging.