Optimize collective intelligence, Hypermind use case

InterContinental Hotels Group (IHG) was looking for innovative new ways to improve the customer experience. They wanted to tap into the enormous creative potential of their employees. Our Hypermind Consulting team designed a prediction market to help them harvest and prioritize ideas from their 1,000-person IT staff. Branded internally as “The Idea Pageant”, this prediction market encouraged participants to anonymously submit ideas and then to bet on the best and worst ideas. In the end, pageant winners were engineers, analysts and contractors—not managers. The resulting strategy and portfolio of initiatives continued to drive performance more than two years later.

Our customer—a leader in global food distribution—knows that their success is dependent on their ability to anticipate the evolution of market prices. Recently, however, after many years of relative stability, the market prices of their products suddenly increased due to exploding demand in emerging countries … and their forecasting models based on backward-looking historical data were immediately useless. The Hypermind Consulting team designed a prediction market to gather the informed guesswork of the company’s many traders and other stakeholders and then consolidate the data into hard forecasts. Ours was one of five forecasting methods that the company used to compute their final forecasts. At first, Hypermind’s prediction market weighed just 20% of the final forecast, but its results proved to be so much more accurate than the other four methods that the company bumped its weight up to 50%.

The United States Air Force’s Rapid Reaction Innovation Team (RRIT) used our Hypermind Cloud Solutions to collect and prioritize the ideas of 2,000 junior engineers at its research lab. Rather than relying solely on the ideas proposed by industrial partners or senior staff, the Air Force encouraged its youngest geeks to propose innovative projects. The result? Better drone cameras and a much more sensitive new radar system.

Molex, an American company specialized in connectivity, has been using Hypermind Cloud Solutions since 2012. Every 18 months, Molex harvests new product ideas from more than 5,000 engineers across their laboratories and factories in the United States, China, Japan, and South Korea. By involving everyone in the ideation process, Molex helps ensures the entire company is on board with new product strategies going forward.

One of our customers—a French media group—called on Hypermind Consulting to help defuse a potentially problematic situation. Management at the company wanted to reorganize schedules at call centers to better respond to customer demands. The quality of life of some employees could be quite negatively impacted. Union representatives blocked the change, and things seemed to be at a standstill. We helped the company set up a consultation of the several hundred employees, not just on schedules but also on a variety of other hot topics. Under the guise of anonymity, everyone could propose ideas, everyone could discuss them, and everyone could bet on each idea’s chances of being accepted by the managers. It was not a question of voting for one’s own preferences or claims, but instead of wagering on what would be acceptable to all. The blocking points quickly dissolved in the face of supercollective creativity. A successful negotiation could finally begin.

Private banking has been around for centuries, and in some ways, it is starting to show its age.
In 2015, a large international private banking firm started to wonder how to stay relevant. An internal brainstorming session generated twenty ideas for new services, new products, new types of interactions with customers. But doubts persisted: Were the ideas any good? Had they thought of all the good ideas? We helped this customer find answers (and great ideas) by submitting the question to 200 expert forecasters from our Hypermind Panel. With the supercollective intelligence our panel provided, our customer quickly eliminated the wobbly ideas and identified the best ideas from the original list—and went home with a long list of great new ideas they hadn’t had internally.

Improving the quality of geopolitical forecasts is part of the core mission of the CIA and other United States intelligence agencies. That’s why the Intelligence Advanced Research Projects Activity (IARPA) funds large-scale research into crowd-based forecasting. From 2012 to 2015, Hypermind Lab’s experts in prediction markets were recruited into IARPA’s ACE research program by a team led by Philip Tetlock (University of Pennsylvania) and Don Moore (UC Berkeley). According to Jason Matheny, ACE program manager and later IARPA Director, this research led to “the largest improvement in judgmental forecasting accuracy observed in the literature.”

In 2017, IARPA launched another multi-year research program called Hybrid Forecasting Competition (HFC) aimed at understanding how to best combine human and computer forecasting abilities. Hypermind Lab’s scientists and developers are again at the forefront of that essential research, as part of a team led by Raytheon BBN Technologies.