Thomas Davenport: Overcoming decision deficit disorder
Thomas Davenport, a professor of Information Technology and Management at Babson College and author of "Competing on Analytics: The Science of Winning," says his research shows that real-life organizations, like Dilbert's company, often don't use good decision processes. Speaking at the Center for Advancing Business through IT (CABIT) Showcase Series, Davenport said that "decision deficit disorder" leads to train wrecks in organizations. But how does an organization get from making decisions badly to making decisions well? To answer that question, Davenport interviewed leaders at 57 organizations and asked how they made decisions. Based on that research, he created a four-step system for making decisions better.
There is perhaps no one better at capturing corporate ineptitude than Scott Adams and his white-collar working man's hero, Dilbert. So it was fitting that Thomas Davenport opened his presentation — on the consequences of bad decision making and how to make decisions better — with a Dilbert cartoon.
Pointy-haired boss: "All important decision-makers in the company are in this room. No little people are allowed because we'll be making important strategy decisions. First, let's make decisions about Project Opal. Does anyone know what the project is or what we need to decide?"
Manager 1: "My executive intuition tells me that we should cut the budget by 10 percent."
Manager 2: "I think Opal is one of your projects. It's named after your daughter."
Manager 1: "Wait ... a new intuition is coming in now ... It says to increase the budget."
Dilbert to pointy-haired boss as he exits the meeting: "Why are those meetings secret?"
Pointy-haired boss: "You don't want to know."
Speaking at the Center for Advancing Business through IT Showcase Series, Davenport shared the results of his recent study of how decisions are made.
Bad decision making processes and bad decisions
During the course of his study, Davenport, who is a professor of Information Technology and Management at Babson College, found that real-life organizations, like Dilbert's company, "don't use good decision processes in many cases."
He most recently observed what he calls decision deficit disorder during a visit to the National Security Agency (NSA). NSA, charged with collecting, processing, analyzing, and disseminating intelligence information, takes in a huge amount of data, including transcripts of communications into and out of the country.
The agency processes about half of the data it captures, NSA managers told Davenport. And it analyzes half of that. "Yet a recent report found that the majority of analysts were not very quantitative in their research; in fact, they preferred to act and think in experiential and intuitive sorts of ways. In other words, they are not very analytical analysts," Davenport said.
Then after the analyses are synthesized, they're delivered to policymakers — who are also "not that analytically-focused in their decision making," said Davenport. "So you have a vast amount of data going into this funnel and being analyzed, but at the other end of the funnel decisions are made on other criteria."
Davenport called the NSA example "sort of extreme" because most organizations don't deal with the huge amount of quantitative inputs that the NSA does. But, he said "there are a fair number of other organizations" that don't have effective decision making processes in place. And dysfunctional decision making processes, Davenport said, often lead to bad decisions.
The negative consequences of making bad decisions — or not making better decisions — are huge. Davenport ticked off a long list of what he called decision train wrecks: Going deep and leveraged into real estate loans at Lehman Brothers. Approving sub-prime mortgages with scant documentation. Rejecting Microsoft's purchase offer at Yahoo. Invading Iraq to find and destroy weapons of mass destruction. Escalating the Vietnam War. Invading Cuba at the Bay of Pigs. Launching the Challenger. Not rescuing the Columbia.
Those negative consequences should motivate organizations to improve the way they make decisions, said Davenport. But how does an organization get from making decisions badly to making decisions well?
How to systematically make decisions better
To answer that question, Davenport interviewed leaders at 57 organizations and asked about how they made decisions. Based on that research, he created a four-step system for making decisions better — and making better decisions.
The first step is to "identify which five-to-ten decisions are the most critical to your strategy." An organization should ask: What businesses should we be in? What is the value proposition for our customers? How do we gain share and grow? Then, "identify the two or three operational decisions that are critical to the execution of each strategy decision."
Identifying the most critical decisions will allow an organization to focus its reform efforts there. "Without some prioritization, all decisions will be treated as equal — which probably means that the important ones won't be analyzed with care," writes Davenport in his recent article, "Make Better Decisions," which appeared in the Harvard Business Review.
At a pharmaceutical company that Davenport studied, the identification of the most critical decisions was paramount. "When it costs $1.2 billion, on average, to take a drug to market, deciding whether or not to move forward with a new drug is a substantial operational decision," he said.
Yet no one was responsible for decision making across the life cycle of a product — the person making the go-to-market decision hadn't been involved in the drug's development since the beginning. "To bridge that gap, the company created super-product managers whose job it was to keep track of those major decisions for a drug product all the way through the process," Davenport said.
The second step in making decisions better is to inventory key decisions — to "assess the factors that go into each", writes Davenport. In this phase, the organization should ask: Who makes the decision? How often is it made? How long does it take? What decision making process is currently being used and does it work?
At Educational Testing Service, another of Davenport's 57 organizations, "decision makers often lacked information" and "it was unclear who played what roles when a decision was being made," Davenport writes.
To resolve those barriers, ETS "introduced a centralized deliberative body to make decisions about new offerings, developed forms that required new metrics for and information about each proposal, and established standards for what constituted strong evidence that the offering fit with ETS strategy and likely market demand." The improved decision making process, Davenport writes, has "clearly resulted in fewer bad product-launch decisions."
The third step to improve decision making is to "decide how to decide" — to "design the roles, processes, systems, and behaviors your organization should be using" to make decisions. For every identified and inventoried decision, the organization must decide what decision making process it will use and who will participate. Then, what information will be used as input into the decision making process? What technology will be used for analysis?
Decision interventions and institutionalization
"Deciding how to decide" leads to what Davenport calls "interventions" — or changes in the way an organization makes decisions. Some interventions are behavioral, like changing how a decision is communicated; others are organizational, like adopting knowledge-sharing approaches; others are technological, like building new data repositories. But it's not a one-shot deal; the companies Davenport studied conducted an average of 5.2 interventions in order to generate better decision making.
Stanley Works' intervention was in the area of pricing. In 2003, they established the Pricing Center of Excellence. The company developed different pricing methodologies; implemented new pricing optimization software; developed an offshore capability to gather competitive pricing data; and automated some promotions pricing. The company built in new pricing responsibilities at the business unit manager level; and to motivate people they hold performance calls with the different business unit executives and the CEO.
The results, Davenport said, are impressive: 6 percent improvement in gross margin, "which is a lot of money on the bottom line." Stanley Works' success, Davenport said, is illustrative of the fact that "if you want to be successful at something like pricing, it's not just putting in pricing optimization software; it's not just creating a pricing center of excellence, it's a lot of those changes working in concert."
The final step in making decisions better is to institutionalize those improved decision making processes. Analysts are an important part of that institutionalized process. But, said Davenport, "The analyst's role is really starting to change."
Now, the role of the analyst involves telling a story with data. "Human beings like stories. But stories with data are better than stories without data. They're more likely to be correct, for one thing. In general, if you want to persuade managers to act, you're going to have to learn how to tell a story with data," Davenport said.
Another key quality organizations have begun to look for in their analysts is courage. "You want the kind of culture where an analyst can say to a manager, 'the data don't support that decision.' Intel is famous for that culture of argumentation based on evidence." And, where analysts were once exclusively number-crunchers, they're now more often expected to understand business operations and strategy, too. They're expected to help "frame the decision" within the context of the organization's strategy.
Yet no matter what it does to indentify, inventory, intervene, and institutionalize better decision making, no organization will always make the right decisions. And hindsight will always be 20/20. "I like to say that a good decision is one that is made through a good process and has a strong likelihood of a positive outcome," said Davenport.
There will always be pointy-haired bosses and managers with their Opal projects and errant intuition. Yet as Thomas Davenport's most recent research shows, the companies that have learned how to create better decision making processes, to overcome decision deficit disorder and make better decisions, have benefited where it counts most — on the bottom line.
Bottom line:
- Many organizations don't use good decision making processes. And it shows in a long list of bad decisions that have had seriously negative consequences.
- The first step to systematically make decisions better is to identify decisions that are critical to the execution of business strategy. Then, identify critical operational decisions.
- The second step is to inventory key decisions — to understand the factors that affect each decision.
- The third step to improve decision making is to "decide how to decide." That leads to what Davenport calls "interventions" — changes in the way an organization makes decisions.
- The final step is to institutionalize improved decision making processes.
- Analysts, who are part of the institutionalization of decision making, should be able to "tell a story with data," have the courage to stand firm when a manager's intuition is not backed up by evidence, and understand business operations and strategy.
Latest news
- A new chapter for Sun Devil Athletics
Sun Devil Athletics Director and two-time W. P.
- ASU AI expert recognized for impact in information systems research
Pei-yu Chen was honored for her contributions to the Management Science Journal.
- Data analytics expert receives prestigious award for dedication to information systems community
World-renowned artificial intelligence and data analytics expert Olivia Liu Sheng was honored…