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.

Comments (1)

Tell me how we’re connected?

April 2, 2008 at 4:35 pm · Filed under Collaboration, Distributed Work, KM, Reviews, Talent Management, social networks

I had a recent conversation with Kate Ehrlich on the topic of expertise location, following on my reading of a blog, “Googlizing Knowledge Management at IBM” by AppGap colleague Bill Ives. Kate has been publishing research in this area for quite some time, and was part of the SmallBlue team whose work became IBM’s Atlas. Bill’s blog also mentioned the IBM Human Capital Study and IBM’s related consulting services. The key point of the study (for me) was the emphasis that companies who rated themselves “Very capable” of adapting to changing business conditions placed on the following capabilities:

• Predicting required skills and their future availability over the next three to five years
• Identifying and locating individuals with specific expertise
• Fostering an environment that allows people to collaborate across organizational boundaries

Let’s look at these last two. Not only is it important to identify people with specific expertise, but it’s also important to foster an environment in which people can collaborate. The ability to collaborate has a lot to do with the capability for connection and the ease with which that connection may be made.

What this means to me is that when I search for something, inside or outside a company, I want to find the best “stuff” but I also want to know which people are the major contributors to the best stuff, and furthermore I want to know how I am connected to them. That’s what Atlas, Metasight, and the newly launched SONAR platform from Trampoline Systems is designed to do. (Trampoline’s product suite includes Metascope, software for social network analysis, and Flightdeck, an “organisational intelligence tool”. I’ll blog further about each of these anon.)

SONAR is user-centric. My dashboard shows my profile (which I can modify) and the themes that occur in the documents I write, the blogs and wikis I post to, and in my emails. A natural language processing (NLP) engine extracts themes from content in the intranet and attaches these to individual profiles. So when I search, I get a list of relevant themes and get to pick which one is most applicable to what I am looking for. When I’m satisfied with that, SONAR shows me (on my dashboard), brief profiles of the people in my organization who are most associated with that theme. Then, I can click on “Explore” and see how all the people tagged with that theme are related. I can also see if any of these people share other interests or themes with me.

Now I can figure out, when I see the work associated with what I am looking for, how various people who might be assumed to be able to help me, are related (to each other and to me). This is a new form of metadata, the metadata of relationships that is so intriguing to me. Our modern lives can feel very fragmented, especially in organizations where people are working on multiple projects and multiple teams at the same time. Each project, team, meeting, has a context that needs to be discoverable and made known. Visual thinking is very important (as in Shiv’s post of a few days ago) but visual prompts and cues such as we are not starting to see in software can also inspire new thought.

I was very excited when Trampoline’s CEO Charles Armstrong contacted me a few weeks ago to tell me about these great developments; I had last seen him over a year ago and the scope of these products had not yet been fully envisioned. I must disclose that I am keen enough to get my hands on these products that I might enter into some business arrangement with Trampoline or a partner.

Comments (5)