If you aren't familiar with the term, "Moneyball," it originates with the Oakland A's general manager Billy Beane's (and his mentor Sandy Alderson). The Moneyball method helped Oakland build a playoff baseball team on a smaller budget by relying on statistical analysis to acquire new players. When it came to baseball talent, Oakland knew it could never financially go head to head with the biggest teams in major league baseball, so the team started using an unheard of method in baseball, the Sabermetric principles.
This research and analytics based approach helped the A's identify undervalued athletes in the competitive MLB talent pool (as well decide which high school and college players to draft). Using this method the A's were able to reach the playoffs three consecutive years. Catalyst IT is applying similar principles to the IT hiring process to help companies build a better team of IT professionals.
In 2001 Michael Rosenbaum, president and founder of Catalyst IT services began his journey with the goal of recruiting IT talent based on metrics rather than traditional hiring methods which include resumes, phone interviews and in-person interviews.
According to Rosenbaum, traditional hiring methods are biased and rely too heavily on the interviewer's perspective. "Resumes and interviews have never been a great way to figure out whether or not someone is going to be good in a job or role," says Rosenbaum. You wouldn't want to trust your company's bottom line to a hunch, so why would you hire people on your team that way?
Although Catalyst IT keeps the actual signals and data its uses confidential, Rosenbaum shares some insight into what he's learned from 10 years of using a Moneyball-like model for hiring IT workers and building Agile development teams.
Hiring practices meet big data
"This method enables us to look at performance metrics instead of simply credentials, which haven't really been able to predict whether someone will deliver," says Rosenbaum. The idea behind this Moneyball approach is to marry big data with the hiring process.
Using massive amounts of data across all the persons in their organization, Catalyst says it can predict with some certainty who will be the high performers on any given project. They use data-points such as how many functions/projects can a team complete in two-week window, how many can an individual complete in a two-week window, defect and rework rates, QA metrics, social networking data points, how they interact when applying online and more. "We are typically looking at a couple thousand data points," says Rosenbaum.
Rosenbaum says he plans to hire 150 people at Catalyst over the next year and estimates the company will look at more than 10,000 applicants before choosing ones that they feel are not only great employees but also that they fit into the culture of the company they will be working with.
The process is almost completely automated, says Rosenbaum and starts with an online application that asks for some basic information. Using algorithms they choose candidates for the next portion, which is a much more detailed online application that takes a few hours to complete. Applicants are being monitored and judged via resume data, keystrokes, time on page, public domain data and other proprietary signals.