“I turn 58 soon, and I’m starting to realize that I may not live long enough to finish many of my great life projects. So I want to try to tempt younger folks to continue them. Hence this call to adventure.”
“Adventure One” encourages “sharp folks willing to get their hands dirty dealing with the complexities of real organizations” to “remake collective decision-making via decision markets”, aka prediction markets.
The last time I worked at a big, public software company, I wished desperately that we used markets internally to make less blunders. I wish that prediction markets were a fundamental part of our society. I wish we used them for science.
The idea of using prediction markets within real-world organisations has always been filed in my mind under “smarter, harder-working people than you have got this covered”. But it’s been almost a decade now, and I still haven’t met anyone who’s worked at a firm that uses systems like these.
So today I’d like to examine the extent to which other people have got it covered, by doing a quick survey of businesses working on this right now.
Founded over a decade ago and with Robin Hanson as the chief scientist, they’ve raised almost $6m and name GE, Best Buy and Motorola as clients. Their usage metrics exceeded my expectations, although I’m curious about what portion of the predictions were made by employees internally vs the public:
“Motorola launched a companywide initiative to solicit ideas and innovations from its employees. In response to the invitation, Motorola’s idea review board was overwhelmed by the 15,000 ideas the system produced.”
This is an impressive rate of engagement, given that Motorola has only 40,000 employees.
The case studies for Best Buy and GE were also based on prediction markets for their employees. Neither specifies when the studies were undertaken, or if either corporation is still a client. The Motorola one seems to have taken place about ten years ago.
Consensus Point talk about their “platform” called Huunu, but the only substantial things I can find on it are what seem to be mockups by a designer:
Based on their employee lineup and website copy (and the fact that they run prediction markets for customers, not just employees), it looks like they’re focusing on market research specifically rather than corporate decision-making in general (e.g. validating new product ideas rather than answering questions such as “will this project be over budget?” or “will this project be delivered on time?”).
I wonder how much of this is due to company politics: it could be embarrassing for executives to find out that their underlings knew all along that an already-underway project is a white elephant. And few suit-wearing power-wielders like being told that their decision-making is suboptimal.
Consensus Point refreshed their website five months ago, and their blog is fairly active, but it’s unclear how successful they are right now — I couldn’t find any information about who their current clients are, or whether they have ambitions to turn Huunu into something more than a vehicle for consulting work.
Still, I was pleasantly surprised to see how well they’ve done. Their advantages over the other players come down to their decade-plus of institutional knowledge and Robin Hanson’s brand.
It looks like they’re trying to tackle corporate governance in general rather than starting with market research:
And it’s not just screenshots, as this case study of their work with an oil company (possibly Shell, based on their client roster) shows, not to mention this recent blog post from the CTO about how prediction markets can be a more accurate, timely “smell test” of a CEO’s competence and integrity.
I’m a big fan of how easy it is to play with their technology:
Initially I thought that AlphaCast was just a slightly fancier version of my old friend PredictionBook (which lets you say how likely you think something is, but not bet on it), but as Adam Siegel kindly pointed out, that’s just a nice interface on top of a sophisticated market backend:
I’m curious about what this notice refers to, and how it differs from their current offerings:
I love how well they’re pitching the idea of prediction markets — everything from their strong calls to action to the sleek website design to the tech itself. If I could invest in only one of these companies, it’d be Cultivate.
Lumenogic’s website is a little dorky, and initially I wasn’t sure if they were still alive, but a quick Google News search put my doubts to rest. And their case studies show that they’re not just paying lip service to the idea of prediction markets:
“The ideas were submitted anonymously, with a description and the benefit to customers and company. For the betting game, the participants were given virtual tokens, each receiving 10 green tokens to be placed on the best ideas and three red for bad ideas.”
I’m stunned by their incredible client roster:
The case study above describes an “Idea Pageant” run for InterContinental. There’s a good write-up in the NY Times about it from 2008. I also enjoyed this Bloomberg write-up of the work Luminogic did with the Air Force a few years ago. Interestingly, it briefly mentions their “approximately half-dozen competitors in the U.S”. I wonder how many of those are flying under the radar, or who have since folded? Please let me know of companies I’ve missed in my roundup.
On a side note, check out this screenshot of their case studies page:
They could stand to learn something from the marketing / branding / design folks at Cultivate Labs. From the looks of it, Luminogic isn’t having any trouble racking up clients, but I get the feeling that crafting beloved products isn’t something they’ve spent a lot of time doing.
No one’s found a way to scale this yet
So, it looks like there are a few competent firms flourishing in this space. One thing I can definitely see is that there’s no off-the-shelf tool yet. The above companies are making money as consultants, AFAICT. One day this will be a SaaS product, where corporations can sign up and start using it straight away without expensive consulting and training fees.
The current state of the field raises my hopes of starting a related business here in Australia, where (as far as I can tell) no one is making waves like any of the above three companies.
This is partly a technological challenge, but mainly an organisational one. It’s a big change, and right now only the most gutsy, smart and well-informed leaders (and teams) will start using these systems without a lot of hand-holding, aka change management, aka extremely long sales cycles.
That will change as the older generation dies out, as the product UX is refined, and as these ideas come into mainstream consciousness through ever more books, articles and success stories.
Max Planck famously said:
“A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.”
Luckily corporate prediction markets don’t need to be “scientific truths” in order for savvy businesses to start using them.