Dot Voting and the Four “R”s: Prediction Markets, Part 2

April 28, 2008 at 3:20 pm · Filed under Change Management, Collective intelligence, Enterprise 2.0, Prediction markets, Web 2.0

In my first post on this topic, I looked at how prediction markets can work like a stock market — buying and trading “shares” that reflect a specific value, such as a ship date, sales forecast, and so on.  Another popular way to use prediction market technologies is an opinion forecast. Employees (or members of a community) can vote on the likelihood that an event will occur or the positive or negative impact of an event.   For instance, one person might enter a value range that indicates “I think there is a 25% to 45% chance that this will happen,” another person might think the probability is 40-50%, another may use wider or narrower ranges, etc.  In this variant of a prediction market, the “winner” is the person who comes closest to the final calculated aggregate percentage within the narrowest margin. For example, if the collective intelligence says the answer is 45%, then the user with the narrower range (40% to 50%) gets more points.

I happened to be in a room with a pair of scenario planning experts and a leadership development consultant during this informal discussion with Maurice Balick when we noticed the power of seeing others’ predictions. We’ve all used facilitation methods that rely on “dot voting” to display the collective wisdom in a room.  Dot voting is a great technique to use when there are multiple courses of action identified during a workshop or meeting and it’s important to narrow down the choices or see where people’s heads are at. Everyone in the room gets a certain number of colored adhesive dots and can use these as votes on particular topics or items. As dots accumulate on one or two of a long list, it’s easy to see where the room is, collectively.

Balick had previously explained that Newsfutures uses four “R”s as design principles in setting up prediction markets. These must be designed into the environment for successful adoption of prediction markets:

• Relevance – it has to matter to the company or participants
• Rewards – appropriate rewards (monetary rewards, stickers, tee-shirts) need to be in place
• Recognition – people who are the most successful should have “bragging rights”
• Relationship – the market must engender conversation

“Aha!” I said again (but only to myself this time). It comes back to relationships and conversations. While the specific value of the knowledge created by the collective wisdom provides value data to management (who must ultimately decide and act), the process of participating in this medium sparks conversations: it’s about getting people to see not just where they agree with others, but also to see that there are a range of positions possible. By seeking the outliers in estimating and rating, there is the opportunity for the isolated expert or group who may have special knowledge to be recognized and listened to.

When we facilitate a dot-voting session, it’s always important to understand the minority view. The workplace of the future is inclusive; tools like prediction markets  can leverage the necessary diversity.

Employee Participation in the Flow: Prediction Markets, Part 1

April 20, 2008 at 12:27 pm · Filed under Collective intelligence, Prediction markets, Web 2.0, social networks

It’s been nearly four years since Time Magazine published “The End of Management?”, an article about the emergence of prediction markets. It’s been about that long since I heard Bernardo Huberman of HP labs talk at a Complexity and Social Networks Symposium directed by David Lazer. Also about that long since I was introduced to Art Hutchinson, whose blog, Mapping Strategy, includes many entries on prediction markets.

More recently, fellow blogger Jenny Ambrozek and her collaborator Vicky Axelrod published an article in Inside Knowledge Magazine, “Co-creating an organisation’s future“, McKinsey published “The promise of prediction markets: A roundtable,” and the New York Times writes, “Betting to Improve the Odds.”

Here on TheAppGap, prediction markets creep into our conversations but we haven’t brought them to the forefront: Jenny Ambrozek has speculates on how prediction markets might bring more diversity of opinions to geographically dispersed organizations. Jon Husband on Gaming and the Workplace of the Future suggests that video game technologies will enter the workplace; I see prediction markets as perhaps (though Jon may dispute me) an aspect of bringing gaming into the workplace.

All these weak signals got very loud last week, after I spent a morning (hosted by the aforementioned Art Hutchinson) with Maurice Balick of NewsFutures, and a number of concepts started resonating strongly. I can’t cover them all in one blog so am starting a series on some of the ideas that started meshing.

The basics of prediction markets are well covered in the various articles referenced above, and so I’ll just do a brief summary to kick off this series.

In its pure form, a prediction market works like a stock market (people buy and sell shares) in which the stock is a bet, which can be:

• An opinion about the probability of a particular event’s occurrence at a specific time. Will this product ship on its scheduled date?
• An opinion about a specific value of a forecast item. What are the estimated sales of a product? What should we price this product at?
• An opinion about the viability of a new product. What is the likelihood that this product will succeed in the marketplace?

The software applications that manage these markets inside companies tap into the “wisdom of crowds,” (as so beautifully exposited by James Surowiecki in his book by that name). Research and experience with these markets has shown that the aggregated opinion of employees (and sometimes of customers) is almost always more accurate than the forecasts of experts or senior managers. It enables those with marginal or distance voices in the organization to be heard, it leverages a potentially vast diversity of opinions and can shape the way that a company thinks. Note that these tools do not obviate the role of management in decision-making, but rather provide management with augmented intelligence to inform their decisions.

One of the notions that intrigues me so much about these applications (and there are more than I’ve suggested above) is that it enhances the potential for employee engagement and participation. John Seeley Brown has said that Web 2.0 is “a profoundly participatory medium.” Companies like Google and Best Buy have integrated prediction markets into their business processes, and many more companies are introducing them.

Yet, the AIIM Market IQ on Enterprise 2.0 shows that less than one third of survey respondents understand prediction markets, while the rest reported to be somewhat familiar (19%), vaguely familiar (22%) or having no idea (27%). The signals will continue to get stronger, and I think must, for these applications, brought into the enterprise, make for potentially profound employee participation and can perhaps bridge the social gaps between employees and management.