Follow Datanami:
April 20, 2020

In Pursuit of Citizen Data Scientists, Not Unicorns


As the CIO of a $26-billion manufacturer, Gary Cantrell had the will and the means to hire data scientists. He had plenty of data science problems to tackle at Jabil, which manufactures electronic devices on behalf of 300 clients at more than 100 facilities around the world. The problem was, there were no data scientists to be found.

“For the longest period, I was convinced there were only three data scientists in the world, and they just moved around from company to company, getting more and more money, because you couldn’t find these folks,” says Cantrell, who is also the senior vice president of IT at Jabil. “That’s what kicked us off on this program.”

For the past three years, Jabil (pronounced “JAY-bill”) has run almost 200 employees through a four-month course. The Jabil employees enter the course as engineers, analysts, or other business-oriented experts, and they exit as citizen data scientists, ready to tackle data science challenges for the 200,000-person firm.

The program has been a success for Jabil, not only because full-fledged data scientists are about as elusive as unicorns, but also because the educational foundation is built within the context of Jabil’s business and its data.

“The issue we struggled with on the font-end is, if you hire a data scientist, they can be really expert on data and they can be really expert on data analytics,” Cantrell tells Datanami. “What they’re typically not expert on, unless you just happen to get the right person, is understanding the data in the context of your specific business.”

Gary Cantrell is the CIO and SVP of IT at Jabil

Six Sigma for Data

Cantrell compares the Citizen Data Science course with Six Sigma, the manufacturing methodology that emerged from Japan in the 1990s. Just as Six Sigma involves challenging experts to solve a particular process challenge, Jabil’s citizen data science course focuses its students’ educational efforts on solving a certain data science problem in the context of Jabil’s business.

“We pull together cohorts, or cross-functional teams, which identify a business problem that they’re trying to resolve,” Cantrell explains. “We go through and try to develop a use case, if you will, or a project case that we’re going after.”

The attendees spend about a week per month on the Citizen Data Science Program. The first week of the program is spent selecting the use case, including figuring out what data is needed to solve the use case, and setting out how they’re going to attack the problem. The second week is the design phase, where the team focuses on visualizing the data, developing insights, and coming up with a hypothesis for how to address the issue.

The third week of the program is spent on data modeling, further analyses, and AB testing, Cantrell says, while the fourth week involves working through what implementation plans, developing summaries of how it can be applied, and developing more business and use cases.

“That gets spread over about a four-month window. There’s a week per month that we focus on those activities, and the balance of the time they go back to office, back to their day jobs,” Cantrell says. “They start applying what they’ve learned as they go through the program, so that by the end of it, they have the full picture of how to take that data and do the modeling, apply the algorithms, and then translate that into a problem they’re trying to solve in the business.”

Business Context

So far, the program has graduated 182 employees who work for the company, primarily in the U.S., Malaysia, and China. In the U.S., Jabil has put the program together with the University of Southern Florida, while in Malaysia and China, it works with the University of Malaysia and the University of Dayton China Institute, respectively.

Cantrell credits the program’s focus on practical working knowledge, as opposed to academic bookwork, as the key to its success. “The staff at USF has been stellar in terms of working with us to create a curriculum and training program that works for working adults, if you will, and hits the heavy academic content in a way that allows more teams to digest it in a way that is quick and meaningful,” he says.

The Citizen Data Science Program hasn’t completely eliminated the need for credentialed data scientists, according to Cantrell. The company has hired a handful of data scientists over the years, and uses them primarily for modeling. But overall, the program has fulfilled much of Jabil’s data science needs, without resorting to endless snipe (er, data scientist) hunts.

“The key for us is applying it to specific business problems or business issues that we can go and get a return for,” Cantrell says. “The data analytics for data analytics sake doesn’t do our business much good. Applying it to a specific problem is where we really get the lift.”

The company has used data science to improve the efficiency and the quality of its manufacturing operations. Several years ago, it instituted an advanced optical inspection (AOI) project that uses machine learning to gauge quality on the factory floor. It also uses machine learning techniques to optimize its demand planning activities. Data also plays a growing role in coordinating with clients to produce less scrap.

Having a strong data culture is an important aspect of data science success (pgraphis/Shutterstock)

Analytics Culture

As more employees enroll in the Citizen Data Science Program, they will tackle additional business-specific projects, with practical results favored over theoretical precepts. But Cantrell’s aim isn’t entirely tactical in nature, as he’s hoping that an embrace of data science will have a long-term strategic impact too.

“We’re starting to get the message across the organization that the data can help us make better decisions,” Cantrell says. “A lot of manufacturing companies are the same way. You have deeply knowledgeable, long-term member of the team that have great gut instinct and all that.  That’s great when you’re on the team. But in the long term, you have to train more of those or you have to develop better processes to make decisions based on data analytics and facts.

“We’re in the stage now where we’re appreciating the need to make these things repeatable, to leverage the data antics to make those decagons, and take a little of the gut instinct out of it,” he continues. “That’s the part that I think is most critical for us in the long run: Getting the culture instilled, if you will, around relying on the data, leveraging the data, and using that to make key decisions. That’s going to be the big benefit that we get out of it.”

Related Items:

The Critical Importance of Data Literacy

A Practical, Hands-On Approach to Machine Learning Education

Q&A with a DataCamp Counselor