The Problem with Averages in Management Decision-Making

In the complex and ever-evolving world of business, decision-making is a crucial skill for managers. Whether it's evaluating a potential merger, allocating resources for research and development, or launching a new product, the quality of these decisions can have a profound impact on an organization's success. While many factors contribute to the effectiveness of managerial decisions, one often overlooked pitfall is the overreliance on averages.

The Averaging Fallacy

Consider a scenario where a manager is tasked with determining the return on investment (ROI) for a strategic business initiative. The manager estimates that this initiative will generate an average of $10 million in free cash flow (FCF) and requires an initial investment of $8 million. Using the standard ROI formula, they calculate the average ROI as follows:

ROI = ($10M - $8M) / ($8M) = 25%

Based on these average figures, the manager might conclude that the ROI is a solid 25%, making the initiative seem like an attractive investment. However, this approach has a fundamental flaw—it assumes that the outcomes are deterministic and that the ROI will always be exactly 25%, which is rarely the case in the real world.

The Reality of Stochastic Uncertainty

In reality, business outcomes are subject to stochastic uncertainty, meaning they are influenced by a range of unpredictable factors. To illustrate this, let's consider three different probability distributions.

1. Uniform Distribution: In this scenario, the manager acknowledges that the ROI could fall anywhere between 8.25% and 11.75%, with each scenario being equally likely. The most likely case will be 25% and there is 0% that the investment turns out terrible (negative ROI) or really great investment (ROI > 100%) with 0% probability.  This distribution reflects uncertainty, however not a very realistic distribution for most business situations.

2. Normal Distribution: In this case, the ROI still has an average of 25%, but it can vary widely, following a bell-shaped curve. Nonetheless, a ROI of 25% is still your most likely outcome. This distribution accounts for the fact that some outcomes are more likely than others, but outcomes from  -38.5% (worst case) and 81.7% (your best case) are still possible. The probability of a bad investment (negative ROI) is very low 2.2% as is the probability of a great investment (ROI > 100%) with 0% probability.

3. Long Tail Distribution: This distribution introduces the possibility of extreme scenarios. While the average ROI remains 25%, the most likely ROI is only 5%. The range of potential outcomes is truly staggering, spanning from a dismal -99% to a remarkable 933%. This wide spectrum of possibilities illustrates the sudden and unexpected extremes that can shape the world of business decision-making. Within this range, there is a substantial 47% probability of facing a negative return on investment (ROI), highlighting the considerable risk associated with such ventures. On the flip side, there exists a 17% chance of experiencing a highly profitable investment, with the ROI exceeding an impressive 100%. These extreme cases underscore the critical importance of acknowledging and preparing for the full range of potential outcomes when making business decisions in an uncertain world. And this is the most realistic but less used kind of distribution for scenarios in business.

The Impact on Decision-Making

When decision-makers rely solely on average figures, they neglect the vast spectrum of possible outcomes and their associated probabilities. In the uniform distribution scenario, the average ROI of 25% may seem reasonable, but there's a significant chance of experiencing outcomes far below or above this figure. Similarly, in the long tail distribution, the potential for catastrophic losses or extraordinary gains cannot be ignored.

In the real world, managerial decisions are influenced by a multitude of variables, each with its own degree of uncertainty. These variables can interact in complex ways, leading to outcomes that do not conform to a simple average. Therefore, making decisions based on averages alone can be misleading and risky.

Start optimizing your management decisions

To navigate the challenges posed by stochastic uncertainty in decision-making, organizations should consider advanced tools and techniques. Leveraging Artificial Intelligence (AI) and decision optimization methods can enable managers to model and simulate various scenarios realistically. By accounting for skewed distributions, correlated variables, and the potential for extreme outcomes, these approaches provide a more comprehensive view of the decision landscape.

In the world of management decision-making, relying on averages can be a dangerous oversimplification of the complex realities that businesses face. Averages fail to capture the full range of possible outcomes and their associated probabilities, leading to misguided decisions. To make more informed and robust decisions, managers must embrace the uncertainty inherent in the business world and leverage advanced tools and techniques that can provide a more accurate and comprehensive understanding of the risks and opportunities at hand. In doing so, they can increase their chances of achieving sustainable success in an uncertain and ever-changing environment.