Customer Story

To succeed, hire the right people.

To find them, look at the data.

A data-driven human resource strategy is the key to increasing employee retention and performance

Septeni

Septeni Holdings

ChallengeAvoid the growth-limiting pitfalls of high employee turnover by eliminating hiring mismatches.
Solution
Predictive modeling has helped identify important trends in employee performance that now inform personnel retention strategy.
ResultsBy matching employees with working groups where they are most likely to succeed, Septeni Holdings reduced overall turnover rates. The company now uses JMP to analyze a wide variety of employee development and performance data.

In the world of human resource (HR) management, data is crucial to understanding employees—from tracking job satisfaction to understanding what motivates individuals to perform at the highest level. Proactive, data-driven strategies are the wave of the future in HR; and Septeni Holdings is real life proof-of-concept. Over the past several years, Septeni has implemented forward-thinking measures to prevent resignations due to employee mismatch. Moreover, the company has improved its hiring processes for recent graduates by projecting future performance with predictive models and screening data.

Proprietary data gives a competitive edge in the labor market

Based in Tokyo, Septeni is a digital advertising and media content firm with a growing overseas presence in the Asia Pacific region. Since its founding, the company has experienced rapid growth with now more than 1,000 employees and a presence on the JASDAQ as of 2011. And though Septeni once experienced growing pains typical of the digital start-up world, the company quickly learned to overcome one of its most high-stakes challenges: hiring and retaining high quality employees.

In the early days at Septeni, turnover was high and mismatched hires were many. Knowing that a company is only as good as its employees, Director Isamu Ueno began searching for a better HR development strategy. He found it in a somewhat unexpected place: Michael Lewis’ Moneyball, a book that chronicles how the Oakland Athletics built a competitive baseball team on a low budget with sabermetrics. Ueno had received the book from his colleague, Group President and CEO Koki Sato. Now intent upon forming their own competitive team of Septeni personnel, the two began exploring ways to let the data guide them toward the best and brightest job candidates.

“We are a start-up, so our hiring budget isn’t very big,” explains Tatsuya Shindo of Septeni’s Human Capital Lab. “We couldn’t compete on investments, so the goal was to conquer the labor market by leveraging proprietary data.” 

Retaining top talent with strategic transfers

Such a strategy was quite a novel idea in the HR space even just a few years ago. But after some initial trial and error, Shindo and his team were able to turn their data into a competitive advantage.

Shindo’s first approach was to create a forecast using Excel spreadsheets. Once it became apparent that Excel was limited in what it could do, however, Shindo turned to a friend who told him about JMP. “I had reached a dead end in forecasting,” Shindo recalls. “Because I wanted to introduce JMP into our workplace right away, I downloaded the trial version and immediately looked into purchasing it. We decided to take action very quickly.”

At the time of a new job candidate’s hire at Septeni, the company’s personnel database logs around 170 data points, from demographic information to the results of interview evaluations to Human Logic Laboratory FFS (Five Factors & Stress) responses. As the employee progresses in his or her career at Septeni, more data is collected. In ten years of employment, an individual record can reach over 800 data points. By using this database to construct predictive models, Shindo is able to better identify trends and modify human resource development policy accordingly. One such model, for example, helps Shindo to predict the likelihood an employee will quit – and to act quickly to improve that employee’s job satisfaction.

When management recognized how repeated verification brought increasingly accurate predictions, this data started to play an increasingly prominent role at the Human Capital Lab. “By transferring several top employees who previously seemed likely to quit, we were able to lower our turnover,” says Shindo. “When you actually manage employees with data, they don’t quit; rather, they continue to work enthusiastically.”

Shindo’s model demonstrates that the likelihood that an employee will quit is correlated with his or her ‘compatibility’ within the workplace. Compatibility, he says, is defined by the relationship between an employee’s temperament (i.e., his or her personality) and the workplace environment (i.e., that of his or her colleagues).

With a transfer, an employee is moved into a new working group – ideally one that is better suited to that particular individual, thereby keeping employees engaged in their work and team responsibilities. Interestingly, Septeni now puts compatibility on equal footing with the company’s other evaluation criteria, conceptualizing the relationship with the equation: Evaluation (Growth) = Personality x Environment [Team + Work].

With the adoption of this approach, the HR group has also changed a great deal. Rather than train employees to have specialized knowledge, Septeni now prefers that employees train themselves through direct on-the-job experience and the mentorship of colleagues. Septeni defines human resource development as “working to scientifically predict and analyze whether employees will acquire quality experiences.” 

Tatsushi Shindo

Data-driven resource management with JMP gives us the evaluation criteria needed to support all our employees as they grow to be more productive personnel. 

Tatsuya Shindo,
Human Capital Lab 

A better workforce begins with data‑driven recruitment

Septeni now uses similar personality-based predictive modeling tactics during the hiring process. Using data collected with an FFS rubric, HR professionals are able to model a prospective employee’s future performance, thus informing recruiters’ decision as to whether a candidate will progress to the next round of screening.

“Ability isn’t the only factor we look for when hiring,” says Shindo. “We screen to see whether candidates are a good match with Septeni’s environment. We want to hire people who will show us what they can do while fully enjoying our work environment. That way, we can prevent employees from quitting when they prove to be ill-suited.” From screening job seekers to placing new employees within the organization, Septeni’s HR group uses JMP to pair new employees with the work environment in which they are most likely to be successful. Their success, Shindo says, ultimately also means the success of the whole company.

“In the future, I would like to create a model that will support any personnel and management decision related to human resources,” says Shindo. With this approach, the company would collect data throughout employees’ careers, as they accumulate experience from the time of their hire through subsequent promotions.

“We hope to grow the quality of our workforce,” Shindo says. “Because we hire employees who have been uniquely matched to our company’s culture, they will naturally produce good results if we provide an atmosphere in which they can optimize their talents. Data-driven resource management with JMP gives us the evaluation criteria needed to support all our employees as they grow to be more productive personnel.” 

The results illustrated in this article are specific to the particular situations, business models, data input and computing environments described herein. Each SAS customer’s experience is unique, based on business and technical variables, and all statements must be considered nontypical. Actual savings, results and performance characteristics will vary depending on individual customer configurations and conditions. SAS does not guarantee or represent that every customer will achieve similar results. The only warranties for SAS products and services are those that are set forth in the express warranty statements in the written agreement for such products and services. Nothing herein should be construed as constituting an additional warranty. Customers have shared their successes with SAS as part of an agreed-upon contractual exchange or project success summarization following a successful implementation of SAS software.

Back to Top