Crowdsourcing ideas is quickly becoming the standard procedure in enterprise innovation. The diversity of ideas and input that large groups of people can provide makes innovation programs more agile and predictable. This makes gathering and selecting ideas faster and more fruitful, but leaves the financial impacts out of the picture. However, it turns out that the crowd is not only great at coming up with good ideas; it can also predict the value of ideas with startling accuracy.
How the Crowd Accurately Predicts Outcomes
Research on the accuracy of crowd predicted outcomes is long, and sometimes a bit unusual. The basis of the question is always fairly similar, though. If you ask a large group of people a quantifiable question, will its answer be more accurate more often than that of an individual expert? The short answer: yes, always.
One of the earliest—and most unusual—examples of proving the “wisdom of crowds” was a study performed by Francis Galton in 1906 at a livestock fair. He asked a group of people to guess the weight of an ox that was up for auction. Galton received 800 guesses from the crowd and found that the average of their answers was only 0.8 percent off of the actual weight. This crowdsourced average was not only very close, but better than that of any individual or singular expert.
More recently, Credit Suisse Strategist Michael Mauboussin ran similar experiments at Columbia Business School, only asking the crowd for guesses for the number of jelly beans in a jar. In his experiments he proved that crowds of sufficient size, diversity, and intelligence will always make better predictions than the individual. Not sometimes—always.
What Makes a Good Crowd?
There are four qualities that make an “intelligent” crowd. First, the crowd must be diverse in that each person should have their own opinion. Secondly, individuals must have some degree of independence, so those around them must not determine their opinions. Thirdly, the crowd must be decentralizedso that people are able to specialize and draw on local knowledge, not one common source. Finally there must be a mechanism for aggregation, which is designed to turn individual judgment into collective decisions.
These may seem like substantial obstacles to overcome, but it does make the business enterprise environment the perfect place for predicting outcomes, given that the right tools are available to aggregate the data and execute on it.
Why it Matters
Most innovation programs can produce a lot of good ideas, but it can be difficult to quantify the potential value of any one idea—let alone the value of the whole innovation pipeline—without the proper mechanisms. Since the crowd is so good at predicting outcomes, it makes sense that they are used to gather quantifiable data for making better decisions in innovation programs. The crowd can fill in this blank in the innovation program by taking this theory of the wisdom of the crowd and applying it to evaluating ideas. When the return on investment of ideas can be predicted, it adds only valuable, quantitative data to the idea selection process. If you know the level of investment required and the potential return of an idea, it makes it much easier to justify decision making. And if the value of all ideas is known, that makes the value of the entire innovation pipeline discoverable as well.
How it Works in Planview IdeaPlace Predictions
This is where the new Planview IdeaPlace Predictions module comes in. Predictions embodies the four qualities required for accurate crowd predictions. Thediversity and decentralization criteria will be met with a large enough group, so the prediction stage should be used with 100+ participants. Theindependence condition is built into the system by the generation of independent samples for the user to vote on, rather than asking the user to input a value (see below). Finally, Predictions, as a mechanism for collecting opinion, and the algorithms underneath satisfy the aggregation condition.
The Predictions Module is used as a part of a larger innovation challenge. Toward the end of the challenge, participants enter the Predictions Phase and they are asked up to three questions regarding the top ideas that have already been selected:
- How much revenue will the idea generate per year?
- What will the cost of implementation be?
- How long will it take to implement?
The answers to these questions are aggregated and a result is calculated using an advanced algorithm. The results are then displayed on leaderboards, and the overall results are also displayed in advanced data visualizations in the form of bubble charts that make it easy to filter the results and see the details.
With this data, organizations can forecast the revenue, time-to-market, and implementation costs associated with an idea. These predictions provide valuable quantitative data that allows organizations to make better business decisions and forecast the ROI of the entire innovation program.