Human Resources | May 1, 2017

From Spreadsheets to Data Science: Cummins Revs Up Its Talent Engine

By: Rob Preston


For a company that prides itself on its world-class engineering and its “hire to develop” culture, Cummins had a decidedly low-tech, inefficient process for identifying its top talent.

Every year Cummins’ HR organization would contact more than 1,800 senior managers worldwide and ask them to assess the leadership potential of their direct reports. And then it would spend several months sending out reminders, chasing down and consolidating spreadsheets, and overseeing a series of “calibration” meetings where the senior managers aligned on who the top talents were in their organizations.

“Then once that was all said and done, it would take us another three weeks just to put some very basic analytics together for our CHRO,” says Fabio Fukuda, director of global HRIS and business intelligence at the Columbus, Indiana-based maker of engines and power-generation systems. “So there you have it. You’re looking at several months of the year, just to execute one—though really important—HR process.”

That one process was representative of a broader HR system problem at Cummins: too many disjointed applications, from four or five different vendors, to manage talent, performance, merit, and other processes; far too much manual work required of HR team members to administer each of those processes; and no easy way to integrate and analyze the data they generated to draw high-value insights.

·         Related: Try Oracle Cloud for free

“We said enough is enough,” Fukuda says. “We needed to create more of an HR technology ecosystem that is easy for the user, puts everything in one place, and automates some of our manual processes.”

From Months to Weeks

The first step was to automate that time-consuming process for assessing leadership potential. Working with the Cummins HR executive team, Fukuda led the global implementation of Oracle Talent Management Cloud’s talent review module, making it much simpler for those 1,800 senior leaders in over 40 countries to appraise the potential of more than 5,000 middle- and upper-management employees.

The data collection part of that self-service process now takes two weeks to complete within one global application, compared with three months on myriad spreadsheets previously. As for the calibration discussions, that part of the process, now a built-in feature of the Oracle HCM Cloud application, takes about four weeks rather than the three months it used to take. The process was completed by 98% of the target executives in year one.

Fukuda’s team also developed several dashboards for Cummins’ CHRO and other senior HR executives, available to them 24/7. The dashboards, built on Oracle Transactional Business Intelligence software, show real-time progress toward completion of the leadership-assessment process, as well as some basic KPIs, such as the number of high-potential employees by line of business, function, and geographical location.

Once the HR executive team saw the power of those dashboards, it made a business case and secured the funding to roll out more advanced analytics, using Oracle Business Intelligence Cloud Service, to Cummins’ top HR leaders. The first wave of that rollout, which combines data from multiple sources and provides sophisticated visualization, is due to be completed in May. The plan is to put some basic metrics in front of the HR leaders, such as how many employees the company is promoting and what the movement of talent is like around the world, and take the analysis from there, Fukuda says.

“Today, our leaders may have a hypothesis—for instance, turnover is high in China—and then they ask for the data,” he says. “Then you have an army of people digging into the data trying to prove or disprove whatever the hypothesis is. That’s good, but it really relies on the leaders’ ability to have those insights. Sometimes people are just busy, and they don’t spend the time to do it. Rather than wait for the insight to come and then they ask for the data, we want to be a step ahead and have the data in front of them 24/7 to drive insights.”

Fukuda, the recent recipient of Oracle’s “It’s All in the Numbers” HCM Cloud Rubies Award, compares this proactive use of data to financial traders making decisions based on information they piece together from their Bloomberg desktop terminals—acting as consumers of data and not just producers. “Now our HR leaders can follow in real time what’s going on in the organization,” he says. “That’s going to lead to more questions, and that’s when we’re going to get into the true analytics.”

Wanted: Curious, Tech-Savvy People

Next up in Cummins’ migration to integrated talent management is the rollout of the performance management, profile management, and workforce compensation modules of Oracle HCM Cloud to some 30,000 employees in more than 50 countries. Those self-service modules, which will eliminate still more rounds of spreadsheet wrangling and administrative work, are due to go live for managers and their people in January 2018, at the beginning of the company’s next fiscal year.

“The value for the user is being able to do it from one platform, not having to go to multiple systems,” Fukuda says. “We’re also building some manager dashboards so they have more visibility into the talent they manage. All of this data in the same system is going to drive a little bit more analytics for the leaders.”

One early outcome of the process automation is that it’s elevating the role of some Cummins HR practitioners from spreadsheet jockeys to advisors and business partners. The company is also looking to cultivate a different mix of HR skills, Fukuda says. For years, the HR team valued people proficient at spreadsheet macros and pivot tables. Now it wants to develop curious HR specialists who are comfortable dealing with higher-level analytics, he says.

“Specifically in the HR analytics space, we want people who aren’t satisfied with what they see, people who are always asking questions—why, why, why—and that makes them dig deeper,” Fukuda says. “They need to be able to make sense of out large data sets, understand data models, understand how complex algorithms work. So they need to be tech savvy. But most important—which is a very difficult combination—they need to understand HR processes.”

Rob Preston is editorial director in Oracle’s Content Central organization.


Rob Preston is editorial director in Oracle's Content Central organization, where he provides insights and analysis on a range of issues important to CIOs and other business technology executives. Rob was previously editor in chief of InformationWeek. You can follow Rob on Twitter at @robpreston.

More about Rob Preston
This is a syndicated post, view the original post here