A pair of recently-published documents from McKinsey & Company and Roland Berger Strategy Consultants suggest that business leaders should spurn strictly quantitative models in favor of alternative predictive tools.
In the January edition of McKinsey Quarterly, the authors of "Have you tested your strategy lately?" offer ten ways that businesspeople can do just that—test their strategies against the firm's famously high standards. From the obvious (No. 1: "Will your strategy beat the market?") to the uncomfortable (No. 8: "Is your strategy tainted by bias?"), these tests run the gamut of business strategy's most oft-broken rules.
The sixth test, "Does your strategy embrace uncertainty?" wonders if business leaders too often base their strategies on specific probabilities, typically calculated by sophisticated predictive models. While not indicting quantitative models specifically (more on that in a moment), McKinsey consultants place their faith in a method that predicts diverse outcomes based on variables that are, crucially, out of a company's control.
Their proprietary method calculates "four levels of uncertainty", ranging from "a reasonably clear view of the future" (level one) to "total ambiguity" (level four). The McKinsey approach is the antithesis of what, the authors suggest, is all too common in modern business settings: companies either assume that they're at level one or level four, and fail to consider the uncertain variables that could alter their projections significantly. It's when companies base their strategies on these blind assumptions that disaster can happen. Companies that consider all the possible outcomes—or, the four levels of uncertainty—stand the best chance of coming through an unforeseen crisis unscathed.