Why Crowd Science Makes Sense: Part I

Filed Under Innovation

Arwen Heredia

February 18, 2014

Using crowd science effectively depends upon a handful of elements: namely, a collaborative crowd, engaged ideators, and established criteria that allow for consistent, repeatable assessments of success. Metrics play a vital role in enterprise innovation programs, and are really the only concrete way to link company efforts to business outcomes. As a result, utilizing crowd science correctly will be a key differentiator for businesses that want to leverage crowdsourcing for innovation.

Good vs. Bad Innovation

It’s all well and good to encourage brainstorming through positivity — after all, one of the biggest reasons employees never voice their ideas is a fear of rejection or failure, and telling people they’ve come up with one or more bad suggestions is a surefire way to silence them in the future. But having the ability to mathematically discern between a ‘good’ idea and a ‘bad’ idea is critical to keeping the innovation pipeline free of bottlenecks and bias. Additionally, using desired outcomes to develop measurements, benchmarks, and qualities of innovation possibilities allows you to shape potential projects before they begin.

For example, in a crowdsourcing environment, employees submit ideas that are then voted on and graduated through the innovation management system. If decision makers determine that a certain number of participants liking or disliking an idea qualifies it as good or bad, predicting the potential of that idea to become a reality becomes a lot easier (and more accurate). Bear in mind that, if you’re basing idea graduation on something as arbitrary (and often opinionated) as good vs. bad, you’ll want fairly dramatic percentages to analyze, which means looking at when the ideas are submitted and who’s interacting with them. That said, if 80% of participants like the idea, it’s a safe bet that it has promise. Similarly, if 65% of the crowd down-votes a suggestion, it’s likely a lot less viable.

The Idea Threshold

This is where things get a bit complex for those of us that don’t intimately understand algorithms and the other, more scientific concepts central to crowd science. That’s because understanding crowd science on a granular level means making sense of all sorts of mathematical formulas, calculations, and variables, particularly when trying to determine whether or not an idea has value.

What you want is to be able to define what data scientists refer to as an “idea threshold.” This is the point (or number of criteria met) that means an innovation has potential and should be supported. According to research, there’s circumstantial evidence indicating that, within the average crowd, innovators (people who come up with great ideas) and discerners (people with a decent track record of supporting innovations that pan out) tend to group around ideas that have high value and potential. This means that, by tracking the behavior of notable participants — as well as considering variables surrounding an idea, like views received, shares, reviews, number of comments, sentiment of comments, number of groups around the idea, etc. — it’s possible to identify the idea threshold that’s appropriate for your business.

Couple that data with calculated info about your participants, and you’re well on your way to better understanding how to mechanize crowdsourcing, grow your innovation efforts, and make projections about future projects. These are just a few reasons why crowd science makes sense for businesses focusing on innovation and driving outcomes — stay tuned for part 2, where we’ll get deeper into the methodology of analyzing your crowdsourcing and innovation programs.

Dig this post? Check out The Science of Innovation: 4 Ways Big Data Informs Innovation Processes.