As I understand it, universal (or even high) participation is not a goal of PMs. The goal is to draw out the subset of employees (or whatever group) that has (a) good information and (b) confidence in the quality of that information.

The more uninformed traders participating, the more noise. With something of value to be gained by smart trading (and something of value [money, time, reputation] to be lost trading blind), a smallish number of the "right" traders will participate and collectively generate meaningful prices.

This leaves open the question of whether PMs are likely to be practical and useful in a wide range of enterprises, and I'm not making any claim that there actually will be adequate participation in any given case. But I do think it's important to see that minority participation is not only acceptable, but desirable.

I saw "Back to the Future" II again recently on tv, you know it is the one in which an almanac is taken back in time. There we have sure bets - 100 percent certainty. However the very fact it was taken back in time added another variable and the actors involved led to new consequences. In many respects the world of the prediction markets is like the sure bet scenario, it is an efficient and certain market within the parameters of the programming - but to take this knowledge outside of that world and use it in the real world - well that is just like taking the almanac into the past. Think of all those games that have variables drawn from life - such as football management, flight simulation - the majority are as good as the state of programming, and are in this severely flawed. Just because someone is good in a flight simulation game, will not give them the ability to fly under Brooklyn Bridge.

To answer your question, I would say that prediction markets are more and more implemented but maybe not in their true format. There are many examples of prediction markets ersatz or evolutions. From Newsfutures Idea Pageant to stimulate and filter innovation in companies to public sites such as blubet and many other ones.
Furthermore, there are studies showing that a prediction market can work even with a rather small number of participants.
What is important is not having a huge population of participants but the volume of trades to ensure fluidity.

Small trader participation rates, while concerning, are not necessarily problematic. Research has shown that effective markets can be established once a market maker function assures liquidity (which is done automatically with many of the available prediction market applications). The modest levels of trial and adoption of this technique by corporations, despite the strong evidence of its' effectiveness, would seem to indicate a problem earlier in the adoption cycle (most never get to the participation rate issue...at least so far). Our experience at BitInsight (www.bitinsight.com), a consulting firm working with enterprise adoption of prediction markets, is that the organizational concerns you raised present a much more challenging issue to widespread use...which is a problem we of course hope to address!

Participation in a prediction market seems to closely mirror participation in other collective wisdom endeavors. Google's numbers seem very high compared to Wikipedia (where it is reported that about 500 people actively manage the content out of millions of potential contributors). However, even with the low numbers there seem to be very beneficial results when the prediction is intended to frame a business decision. From the transcript of the McKinsey roundtable in the Middle East, the comments about the accuracy of the participants in Best Buy's prediction markets is amazing. They were significantly more accurate than the 'experts'. Therefore, I see considerable value in the process even if there are not that many participants. Having realistic expectations from the outset will help to alleviate concerns about lack of participation or adoption.

Your concerns are not exclusive to prediction markets. They are also true for enterprise wikis, Parent Teacher Association activities, and some marriages. A lot of people could contribute, only a few zealots do, but then a sizable number take advantage of the zealots' efforts. The 'gambling' aspect simply entices more people to become engaged in the process.

Tom Davenport
Prediction Markets: Is Anybody Really Predicting?
The participants in prediction markets would have to be pretty highly motivated to keep going back to the site and placing new bets. In any case, as the McKinsey roundtable participants suggest, if you want to run one of these things, you’ve got to devote a lot of attention to marketing them to employees.
How about it—have you tried a prediction market in your company? Did anybody show up?
What you are saying now is the employees ought to be the second-class citizens. I thought that the employees wearing the shirts of the corporation advertise faster then the few outsiders who commute the train hardly noticed by any as all are running for the missed tube. The employees are always visible in the arena rather the buyers come and the neighbours see and ask.
I think the HBR is growing more USA .and not the global. .
I thank you
Firozali A. Mulla MBA PhD
P.O.Box 6044
Dar-Es-Salaam
Tanzania
East Africa

After having done some experiments in our company, my conclusion is that the biggest hurdle for the success of prediction markets is not the number of participants, but it's complexity and abstractness. Try instructing laymen on how buy and sell idea futures with funny money through a website, it sound easier than done. I even built a website (www.groepsintelligentie.nl, in Dutch) as user friendly as possible, and even I have to look up the exact workings when I use it.

I think it is a valuable instrument but with a high threshold, which means that you can use it very well in 'laboratory' situations (gather a group of about 20 people, instruct them and let them trade for an hour or so) but less well in 'continuous' projects.

We have found that prediction or decision markets work best with many rounds of trade, iteratively correcting the stated consensus or average. We have also found that controversial framing of an issue and provocative taunts to participants induces higher levels of participation. Daily email or SMS updates of changes also tout a market, inducing participation. Often well-form visuals depicting the scope, interrelating elements, forces, flows, hierarchy of concepts, etc. really assist "newbies" grasp a prediction or decision more quickly and holistically. Finally, we have found the high-profile public speakers or well-respected thought leaders play an instrumental role in making the prediction or decision market a "game worth playing."

We have proposed markets to our larger clients for the purpose of predicting the effectiveness of marketing campaigns and the constituent elements (regional optimizations of media mixes, guerilla tactics, social media placements, etc). However, it remains a bridge too far; lots of reasons why not to.

So, has anyone conducted or seen prediction or decision markets for marketing programs, media placements, and engagement strategies for online consumers?