A couple of years ago I attended a seminar by University of Toronto professor Avi Goldfarb, who along with UoT colleagues has been conducting research on the economic value of AI. Goldfarb explained that the best way to assess the impact of a new radical technology is to look at how the technology reduces the cost of a widely used function. For example, computers are essentially powerful calculators whose cost of digital operations has been dramatically going down over the past several decades thanks to Moore’s Law. Over the years, we’ve learned to define all kinds of tasks in terms of such digital operations, e.g., inventory management, financial transactions, word processing, photography. Similarly, the economic value of the Internet can be described as reducing the cost of communications and of search, enabling us to find and access all kinds of information, - including text, pictures music and videos.
Viewed through this lens, our AI revolution can be viewed as reducing the cost of predictions. The explosive growth of big data and advanced algorithms like machine learning have enabled us to analyze and extract insights from all that data. Prediction means anticipating what will happen in the future, and given the widespread role of predictions in business, government and our everyday life, AI is already having a major impact on all kinds of human activities. Decision making, the process of identifying and choosing alternatives, will be particularly impacted by advances in AI.
Decisions typically involve two main tasks: predictions and judgement. Judgement is the part of decision making that, unlike prediction, cannot be explicitly described to and performed by a machine and is based on human factors like expertise, intuition, and unconscious feelings. As machine predictions become inexpensive and commonplace, the human judgement that leverages and complements prediction will become more valuable, especially given the increasing complexity of the decisions our institutions are called upon to make.
“It’s the best and worst of times for decision makers,” said McKinsey in a recent article, Untangling your Organization’s Decision Making. “Swelling stockpiles of data, advanced analytics, and intelligent algorithms are providing organizations with powerful new inputs and methods for making all manner of decisions.” On the other hand, “Corporate leaders also are much more aware today than they were 20 years ago of the cognitive biases - anchoring, loss aversion, confirmation bias, and many more - that undermine decision making without our knowing it.”
What accounts for this seeming paradox? Our emerging data and AI revolution holds the promise to significantly augment our judgement and expertise and help us make smarter, more effective decisions. But, explains McKinsey, our growing organizational complexity has clouded decision making accountability. The number of decision makers has risen, making it harder to establish clean lines of responsibility. Moreover, the proliferation of digital communications have led to too many meetings and e-mail threads and too little high-quality dialogue, contributing to disengagement, - “hearing a presentation for the hundredth time”; paralysis, - “stymied by too much data”; anxiety, - “the stakes are too high”; and increasingly poor decisions.
As a first step in understanding decision making, McKinsey categorized the type of decisions being made by most organizations:
- Big-bet decisions have the potential to shape the future of the company. These infrequent, high risk decisions often involve situations with unclear right or wrong choices;
- Cross-cutting decisions are relatively frequent and generally require broad collaboration and an effective decision making process across different groups in the organization;
- Delegated decisions are frequent and low-risk, and can be handled by individuals or working teams with limited input from others in the organization; and
- Ad hoc decisions arise episodically and are generally low-stakes
To better understand the nature of decision making, McKinsey conducted an online survey and garnered responses from over 1,250 participants from a variety of industries in 91 countries. The sample skewed toward upper management, - one-third of respondents were C-level executives and 35% were senior managers. The survey focused on big-bet, cross-cutting and delegated decisions, but didn’t ask about ad hoc decisions because they tend to vary too greatly.
Several weeks ago, McKinsey published the results of its survey. Overall, only 20% of respondents said that their organization excelled at decision making. Respondents from such so-called winning organizations were twice as likely, - 36% compared to 18% for all other organizations, - to say that their recent decision resulted in financial returns of 20% or more. At the same time, only 20% of winning organizations said that their decisions delivered returns of less than 10%, compared to 51% of all other organizations.
In addition, while most organizations make trade-offs between velocity (how fast was the decision made and executed?) and quality (how good was the decision?), the survey showed that the decision-making winners performed well on both velocity and quality, as well as seeing better financial results.
How much time does decision making consume? Quite a bit, according to survey respondents, over half of which said that they spend more than 30% of their working time on decision making, and more than 25% saying that decision making takes a majority of their working time. Not surprisingly, the share of time increases with seniority, with 14% of C-suite respondents saying that they spend over 70% of their time making decisions. However, 61% of respondents said that their decision-making time is not used effectively, include 57% of C-level executives.
The survey also revealed that speed is a bigger decision-making challenge than quality. 57% of respondents said that their organizations make high quality decisions, compared with 48% that said that their organization make high velocity decisions, and 37% that said their organizations made both high quality and high velocity decisions. The results varied by decision type, with big-bet decisions getting the highest markets, - 65% high quality; 55% high velocity; 44% both, - and delegated decisions the lowest: 46% high quality; 40% high velocity; 27% both.
While one might expect that high quality decisions involve more deliberation and should therefore take longer, the results indicated that speed and quality outcomes are highly interrelated. “According to respondents, the organizations that make decisions quickly are twice as likely to make high-quality decisions, compared with the slow decision makers.”
Such a combination of high quality and high velocity is much more common in winning organizations. To understand how they do so, McKinsey analyzed how such winning organizations make decisions compared to all others, and found three foundational best practices:
Make decisions at the right level, often by delegating to lower levels, - a practice 6.8x more likely to be part of a winning company. “This result is closely related to another finding: both high-quality decisions and quick ones are much more common at organizations with fewer reporting layers.”
Focus relentlessly on enterprise-level value, - a practice 2.9x more likely to be found in a winning organization. Overall, only 41% of respondents said that their organizations’ decisions align with corporate strategy and in support of high value projects.
Get commitment from the relevant stakeholders, - a practice 6.8x more likely to be followed by winning companies, which will “build commitment to executing decisions once they are made, especially among the people who are ultimately accountable for a given decision.”
Beyond these foundational practices, winning organizations also demonstrate best practices specific to each decision type:
Bit-get decisions, - 2.3x more likely. “Increase quality of interactions during decision making by exploring alternatives, challenging initial hypotheses, and appointing a devil’s advocate to present counterarguments.”
Cross-cutting decisions, - 4.5x more likely. “Focus on process and coordination of decision meetings to encourage collaboration among individuals and avoid silos.”
Delegated decisions, - 3.9x more likely. “Empower employees to make decisions by creating a strong sense of ownership and accountability and encouraging a greater inclination toward action.”
Comments