Harrah's New Twist on Prediction Markets

In a pilot, casino staff will forecast customer spending, which could help boost revenueand may prompt Harrah's to form more diverse teams

Almost 15 years ago, Lew Platt, the former chief executive of Hewlett-Packard (HPQ), observed that, "if only HP knew what HP knows, we would be three times more productive." This quote became a mantra for the knowledge management movement. Companies invested large sums of money to codify knowledge and connect people.

Unfortunately, many of these initiatives foundered. There were at least two problems. The most valuable knowledge is difficult, if not impossible, to codify. And tools for connecting people run afoul of widespread incentive structures and cultures that strengthen walls rather than foster networks.

There's an even more fundamental issue. Knowledge management, at least as conventionally applied, focuses on tapping into existing knowledge. It spends much less time finding better ways to generate new knowledge.

Harrah's, the casino entertainment company, is beginning to explore how to create commercially useful knowledge by applying a much-hyped knowledge management tool—prediction markets—in a very new context.

The Prediction Market Paradox

Prediction markets are platforms that invite participants to speculate on uncertain future events (such as probability or timing) by buying "shares," with payoffs tied to the eventual outcomes. These hit public attention 20 years ago, and the first corporate prediction market launched a couple of years later. Since then, they have been used in a broad range of industries, most often to estimate the actual release dates for, and likely success of, new products.

Harrah's is setting up a pilot prediction market to forecast customer activity in one of its domestic casino operations. In this market, Harrah's employees will forecast customer spending activity in Harrah's casino based on the employees' front-line experience and perspectives. Harrah's hopes to understand better employees' diverse approaches to generating these forecasts and to explore possible initiatives to increase revenues. The company has also configured "inclusion to innovation" teams to draw on the cognitive diversity among its employees. This taps into the thinking of Scott Page, a professor at the University of Michigan, who published a fascinating book last year, The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools and Societies.

Page argues that diversity should not be confined to differences in identity—for example, race or gender. He makes a powerful case that harnessing differences in how we view the world helps us generate more insight about that world and to be more innovative in our business initiatives. Since the power of prediction markets hinges on effectively tapping into cognitive diversity throughout an organization, Page also argues convincingly that if members of a group do not have enough diversity in their perspectives, prediction markets can actually produce dismal results.

Diverse by Design

Fred Keeton, Harrah's vice-president for external affairs and the chief diversity officer, clearly agrees. "We need to focus more broadly on cognitive diversity, rather than only on identity as a source of diversity," he observed to us in an interview. "This will allow us to configure teams that generate superior economic value by coming up with better solutions to real business problems. In this way, diversity becomes much more than a compliance issue or an abstract ethical value—it becomes a commercial necessity."

The "inclusion to innovation" teams are relatively small teams formed on an ad hoc basis to address specific business challenges within Harrah's more broadly. Team members are selected based on the scope of their relevant experiences as well as on a cognitive diversity profile emerging from answers to a questionnaire designed to reveal differences in approaches to problem-solving.

While still at an early stage in these efforts, Keeton views prediction markets and diverse-by-design business teams as complementary ways to harness the diversity that already exists within Harrah's. Longer term, Keeton sees prediction markets as a way to bring additional pools of diversity to the surface and put together unexpected groups of participants in business teams.

Until now, few of the companies sponsoring successful pilots or tests have deployed prediction markets on a broad or sustained basis. Why not? One explanation is that prediction markets are deeply subversive. After all, lots of midlevel executives are consumed with the task of forecasting. If prediction markets do a better job of it, doesn't that discredit the efforts (and perhaps even the motives) of these executives? But as prediction markets shift their focus toward new knowledge creation, they may become less threatening within corporations.

Bottom Line Lessons for Executives

Knowledge creation trumps knowledge transfer

As Lew Platt pointed out, it isn't easy accessing and connecting all the knowledge that resides within companies. Ironically, the most effective way to improve knowledge transfer may be through more effective knowledge creation, with knowledge transfer as a valuable by-product.

Diversity trumps ability

We all have faith in experts, but superior insight—especially in a rapidly changing world—can often be generated by tapping into distributed knowledge held by large numbers of diverse participants. The Wisdom of Crowds by James Surowiecki popularized this concept, but Page's The Difference provides rigorous analytical support and shows when and why it can go wrong. We have few analytical tools to measure the scope of diversity within our organizations, much less to determine how to harness it within our business teams. Prediction markets can be extremely valuable in making diversity more visible and accessible.

Diversity across firms trumps diversity within companies

We can increase the value of the diversity within companies by identifying and connecting with those from other companies. Suppliers, distributors, product designers…all business partners bring a broad set of experiences, perspectives, and analytic frameworks that augment the abilities of any individual firm. Prediction markets have the potential to tap into this broader diversity.

Power trumps accuracy

We often believe accuracy will prevail on its own. The experience with prediction markets challenges that assumption. Accuracy can be deeply threatening, especially if it is viewed as undermining the authority of managers or specialists. The experience with prediction markets to date suggests that different approaches will be needed if these tools are to be scaled and sustained within most companies. Perhaps prediction markets need to be viewed as part of a broader change-management program that explicitly acknowledges the political dimension of information within organizations.

Before it's here, it's on the Bloomberg Terminal.