One of the underpinnings of crowdsourcing – the reason why it has triggered a thundering herd of start-ups – is the power of the crowd to take on tasks not well suited for computerization. Another fascinating aspect of crowdsourcing is that it can also beat the computer on its home turf of number crunching and data analysis.
My posts on in-house crowdsourcing have focused on technology’s power to connect and harness the creative brain power of virtual communities to form a kind of dynamic contact center.
Business has discovered the crowd, and there are now many new Web 2.0 ventures that rely on anonymous workers, especially for more left-brain oriented assignments. For example, one marketing service I came across uses the crowd to develop and test new product names.
But the crowd can also perform complex right-brain data calculations. The crowd’s data ability was first understood by the late 19th century statistician Francis Galton. His famous wisdom of crowds experiment involved a contest in which the goal was to estimate the weight of an ox.
Long story short, Galton averaged the numeric guesses made by villagers to obtain a result that was uncannily close to the true value. (Aside: by averaging answers, Galton was able to cleverly reduce error variability!)
Scroll ahead to the 1990s. When their own planning models proved unreliable, Hewlett-Packard went to an in-house crowd to predict printer sales. The results of this modern version of Galton’s experiment revealed that employees from many different departments were better at crystal-balling quarterly printer revenue numbers that H-P’s own financial experts.
No one is suggesting that we crowdsource getting account balances from the bank. So when is it appropriate to ask the expert agent, and when do we go to the crowd for answers?
Standard customer-agent interactions work best when the information to solve the request is in one place—a credit card transaction. But the crowd does very well when the data or knowledge about the problem is dispersed among many people.
A sporting goods company that’s trying to decide whether its next sales campaign will feature blue hybrids or racing green touring bicycles may want to ask the crowd for the answer.
