I came across the story of a manager who felt that the best way to encourage desirable behaviours was through reward and humiliation. This encouragement occurred indirectly through what I would describe as “persuasive data”: a table of data went out each week showing the best and worst performing employees. Everyone in the team could see the stats plainly along with the names of coworkers. They were encouraged to make comparisons. This represents an aggressive use of data. From my perspective, this technique has at least two problems: 1) lack of constructive guidance for agents; and 2) an overemphasis of and overdependence on a mono-dimensional metric. For example, the fact that a person isn’t a great builder of widgets today doesn’t mean he or she cannot be a fabulous builder of widgets tomorrow. The fact that a person shows an inability to “build” widgets does not mean that he or she cannot “sell” widgets. Yet, I wonder if the underlying premise of the manager might not be correct at least on a certain level.
There is a principle in human resources that an organization adapts to different external conditions by reorganizing itself internally . For example, if a company wishes to pursue a growth strategy instead of streamlining or austerity, the internal structure that existed to support low-cost production might have to be changed to enable the redirection to growth. Compensation, marketing, and operations might all be affected. I believe that a similar logic follows in relation to data. In order to achieve different organizational outcomes, it is sometimes necessary to reorganize even soft assets such as data. Ideally, there should be not just a new signal regime from the data reporting system; there should also be substantiating discourse to reduce stress, expedite behavioural changes, and encourage team-building. But sometimes, the organization might not “be there” in terms of its supporting resources to allow for caring and thoughtful transitions; or perhaps it doesn’t know how to be caring or thoughtful. One doesn’t necessarily have to be either to succeed – at least in the short-term.
Think about the situation in more abstract terms to appreciate the complicated nature of the undertaking from the perspective of the data specialist. The reconstruction, remodeling, and reorganization of data by the specialist must somehow lead to behavioural modifications among workers resulting in desirable and measurable changes. This blog is the about the extent to which such an endeavour might be possible. I suspect that there are many dozens of ways for a person to measure something and to say something. Also, people are sensitive to being measured and told things. The whole rationale behind making “individual performance” a matter of public record is to exploit that sensitivity. But one should not, as a basis for organizational development, use data to intimidate and diminish workers. As sure as the sun rises in the morning, unless an organization wants to waste time and money constantly hiring and firing people, it will need existing workers to help bring about change; and perhaps this might be an uphill battle when they feel disempowered, exploited, and diminished.
At the same time, if it were possible to influence an organization using something cheap such as tables of data, to me that seems like a worthwhile investment. Consider a more generic scenario that doesn’t involve peer ridicule or embarrassment. The fact of the matter is that the data specialist doesn’t control much of anything except data and deliverables involving data. Ideally this individual or team of individuals doesn’t have much direct connection to most people in the organization. Literally, data specialists might be in an entirely different room, floor, or building – rather insulated from the day-to-day routines of production. (I don’t suggest this is necessarily a good thing. I merely emphasize that a conceptual disconnection exists.) The data specialist doesn’t judge people by their appearance, attire, or social habits. There is a kind of coldness in the work that they do – except for me since I am a naturally warm individual at least in terms of my body temperature. There is no place for pessimism or optimism. There is only data, theory, and ethics.
Think of the organization as a complex system, which probably means that it isn’t static. It is a living and constantly moving creature. I remember a Chinese movie where a character wielded power through the use of seals – slips of paper with writing or symbols on them. Imagine trying to control a bucking organizational behemoth using blessed or enchanted seals. It seems implausible, right? How about by flashing tables of data? Isn’t this in some fashion the underlying concept behind the methods of the manager in my story? The main problem with the concept of a mage or priest “casting spells” isn’t so much the casting part – but rather the fact that he seems to already know in advance what to cast, how to cast, where to cast, and whom to cast it on. It seems intriguing that something possibly complex and intangible such as production data might be constructed and delivered just in such a way to optimize the benefits to the organization: a frowning face can be made into a smiling face with something not all that different from a fortune-cookie message or get-well card.
Not just any data will work. The data must be persuasive. Now, the rationale as I see it isn’t to shock or force behaviour – but rather to be supportive and convincing. I think that an organization has to be seeking the truth in order to embrace it once the truth becomes apparent. The job of the data specialist is to make the truth more accessible in order to build a market for it. What people are seeking and what the market has to offer are not necessarily the same. The objective however isn’t really to sell specific things – not right away – but rather to simply encourage the seeking. Once people are seeking, there is a market that can be fed. The data specialist doesn’t define the truth but helps to encourage the search for it. It is a wonderful thing for an organization to feel empowered enough to search, experiment, and fight. Fight – rather than flight – is a sign of engagement. Avoidance and evasion most certainly lead to market displacement.
I cannot say exactly how the introduction of an aggressive data regime would alter the dynamics of an organization. But I believe that it most certainly can. This being the case at least for me, I consider it of paramount importance that the construction of data be coherent and thoughtful, ethical and respectful, balanced and effective. It goes without saying that the interaction between different complex systems remains an area that requires more research. However, this is a base of knowledge that will likely develop over time over many organizations – perhaps through the efforts of data scientists interested in this type of research. It represents an entirely different kind of management science.
 Richard J. Long, Strategic Compensation in Canada, 4th ed. (Toronto: Nelson Education Ltd., 2010), p. 27-30.