Story:
Unlocking the Hidden Reservoirs of Human Potential: How "Big Oil" and Knack Combined Radically Innovative Big Data Technology and “Big Data Mindset” to Identify Winning Innovators at Scale

Key problem: What makes a successful innovator?  What are the talents, abilities and traits that distinguish high-impact innovators from everyone else?

Key result: Counterintuitively, Knack's data show that the idea submitted at the outset by a would-be innovator did not matter at all; rather, what mattered was the unique behavioral makeup of the person---and, in fact, Knack successfully predicted those people whose ideas turned out to be most successful based exclusively on their behavioral makeup, without having any information about the "quality" of the ideas submmited at the outset.

Underlying Insight: Knack developed its innovative technology to provide a data-driven predictive insight into identifying and unleashing human potential at web scale.  The technology is built on the idea that gameplay---a natural experience that engages, motivates and immerses a person completely--provides an authentic, rich insight into the person's unique behavioral makeup in a way that no other medium or experience can provide.  

Building on this insight, Knack developed engaging mobile games that leverage behavioral neuroscience insights and data logging and data processing infrastructure to capture the rich stream of micro-behavioral data.  

Mining micro-behavior attribute patterns: Knack then mines the resulting gameplay data logged from the games and identifies unique behavioral marker patterns that differentiate people based on their potential for success in innovation (or in any other domain of achievement and success, eg, management and leadership, engineering, medicine, teaching, learning, etc.).

Predictive Models of Human Performance: Knack discovers and quantifies the mathematical relationships between people's "knack footprint" and real-world outcomes.  It provides companies, schools, and other organiztions the ability to create "signatures" that predict high-performance and achievement across a broad spectrum of jobs and fields.  These "Knack Signatures" enable companies and schools to consistently and accurately identify people with the highest potential for success—at large scale.

Knack games: These are accessible across web and mobile platforms.  Unlike traditional methods, Knack games deliver an intrinsically engaging and motivating experience that yields high-quality data resistant to self-report bias.  Knack data generate insights into domains of behavior not accessible using traditional methods. 

Attribute spectrum: These rich data-streams enable the measurement of a wide range of behavioral attributes—from social intelligence and personality traits to creativity, insight, risk-taking, and intelligence—and the discovery of unique patterns of behavioral markers that predict people likely to be successful on the job.

Knack worked with its Big Oil customer to uncover the Knack Signature for High-Impact Innovation and to predict people with the knack for innovation.  The customer identified a group of individuals who had previously submitted ideas; they were invited to play two Knack games.

The games

In Dungeon Scrawl, players navigate a maze and solve puzzles, requiring creative insight, among other talents and traits.

In Wasabi Waiter, players manage a flow of customers in a restaurant setting and have to recognize various social cues, requiring social intelligence and goal-orientation, among other talents and traits.

In Balloon Brigade, players fill with water different types of balloons, and use them strategically to defend their water tower from the onslaught of invading creatures and to save flowers and plants from fire.

200 individuals played two games.  These individuals were selected by the customer.  They had worked at the customer's innovation group.  The customer rated every person as high- or low-impact innovators prior to the test using a 4-point scale.  These ratings were based on the performance of individuals over several years (in some cases, more than 10 years).  These ratings included a measure of both (a) the volume of ideas each person created and (b) the status of the idea---i.e., how far did the person manage to get the idea towards development, testing, and commercialization.  The set of 200 included a distribution of high- and low-impact innovators.

Using several megabytes of gameplay data collected (per person) from 75% of the sample, Knack quantified the Knack Signature that predicts innovation performance.

The signature was tested on the remaining 25% of the sample, which served as a blind test set of previously unseen individuals.  Knack predictions successfully differentiated high- from low-impact innovators in the (previously unseen) test set.