Posted onNovember 17, 2016
The HR world is abuzz with stories about the promises and perils of predictive workforce analytics. For a number of years, organizations have been conducting workforce analytics—using descriptive statistics to summarize workforce events. To these capabilities, many are looking to add a predictive component. Predictive workforce analytics involves using advanced statistical techniques to identify historical workforce patterns in order to predict future behaviors and events. While interest in predictive workforce analytics is strong, most organizations have not yet built predictive workforce analytics capabilities. Luckily, there are some truly innovative organizations doing pioneering predictive work that aspiring organizations can learn from. To help illuminate the work of these rare innovators, APQC partnered with Talent Analytics, Corp., a globally recognized leader in predicting an individual’s performance, pre-hire. As part of a joint-research project, APQC and Talent Analytics, Corp. collected lessons learned from early adopters of predictive workforce analytics.
In summer 2015, the research team conducted structured interviews with workforce analytics leaders from Cargill, Gap, IBM, Johnson Controls, and SAS. The interviews collected information on:
- why the organization conducts predictive workforce analytics,
- how it staffs and structures its workforce analytics capability,
- which data it uses for analyses,
- what the first predictive analytics project entailed, and
- how the results of early predictive analytics work were used.
This research study focused on the practices of organizations that are early adopters of predictive workforce analytics. These organizations stressed, however, that predictive analytics is just one tool in their workforce analytics toolboxes. The practices that these organizations use are therefore key for conducting any workforce analytics project including those that are predictive in nature.
Key Practices for Getting Started
Purpose—Articulate a vision for why your organization is adopting workforce analytics as a business tool.
The early-adopter organizations did not speak about making a large business case before getting started with workforce analytics. However, each did talk about having clearly articulated a vision for why their organization was adopting workforce analytics as a business tool. Essentially, each had answered the broad, long-range question: Why conduct workforce analytics at our organization? And then they crafted specific, short-term workforce analytics plans.
Resources—Secure the specific resources necessary to carry out your organization’s short-term workforce analytics plan.
The early-adopter organizations did not mention making large financial outlays to get started with workforce analytics. For the most part, interviewees did not talk about supporting workforce analytic
with significant investments in technology or new staff. However, they did underscore the importance of securing the specific resources needed to carry out their short-term workforce analytics plans. Each articulated and then located critical workforce analytics skills. Next, they assembled these skills into formal, dedicated workforce analytics groups.
Problems—Select workforce analytics projects in response to business challenges that your organization faces.
The early-adopter organizations do not conduct workforce analytics because they see other organizations doing this work. Instead, their workforce analytics projects arise out of true business problems. Moreover, their analytics projects are only predictive when predictive is the most appropriate method for answering the specific business problem at hand.
Data—Don’t wait for perfect data before getting stated with workforce analytics. Assemble and validate data according to the requirements of your short-term workforce analytics plan.
The early-adopter organizations share a common long-term goal to establish clean, organizationally consistent, and centrally stored workforce data. Some of the early -adopter organizations started work on this goal years back and have made significant progress. Others are still in the beginning stages of data integration. One commonality among them all is the decision not to wait for perfect data before getting started with workforce analytics. Instead, the early-adopter organizations assemble and validate workforce data on a per-project basis.
Education—Educate end users about the basics of workforce analytics.
The early-adopter organizations devote significant time to educating end users about workforce analytics. During projects, they present incremental results and solicit user feedback. Post project, they extensively socialize findings by sharing consumable amounts of information, often in the form of a story. At all times, they provide varying levels of workforce analytics education to HR and other areas of their organizations.
Measures—Measure and share the outcomes of your organization’s workforce analytics efforts.
All of the early-adopter organizations measure the results of their workforce analytics projects. The stories and visuals they create with data promote action, which they closely track as a key measure of analytics success. Any positive outcomes that arise, they deliberately publicize in order to build the business case for continued investment in workforce analytics.
Cargill provides food, agriculture, financial, and industrial products and services across the globe. Cargill has 152,000 employees in 67 countries. For this project, APQC interviewed Michael Crespo, assessment and selection lead, and Jeff Idle, HR business intelligence and analytics lead, at Cargill.
Cargill uses predictive workforce analytics for selecting and assessing individuals in hiring and promotion.
One of its first predictive projects aimed to improve the organization’s ability to select job candidates who have the most potential to develop into high-performing employees. The team created a competency- based hiring assessment that rates how well job candidates fit with characteristics predictive of high performance at Cargill.
Gap is a leading global retailer offering clothing, accessories, and personal care products for men, women, and children under The Gap, Banana Republic, Old Navy, Athleta, and Intermix brands. Gap has more than 140,000 employees and stores in more than 90 countries. For this project, APQC interviewed Anthony Walter, director of workforce analytics, and Andrew LeFevre, senior director HR strategy and workforce analytics, at Gap.
Gap has a workforce analytics center of expertise. One of the Gap’s first predictive workforce analytics projects aimed to identify when critical employees are at risk of leaving the organization. The workforce analytics center of expertise uncovered drivers of turnover at the organization and used these to project turnover for brand leaders.
IBM is a globally integrated technology and consulting company with more than 400,000 employees and operations in more than 170 countries. For this research project, APQC interviewed N. Sadat Shami, manager of IBM’s Center for Engagement and Social Analytics.
Within IBM’s HR function is a predictive social analytics team. One of the team’s first projects was to use social media to get a real-time understanding of employee engagement. The team created a tool called Social Pulse, which uses IBM employees’ social media sentiment to predict if engagement is increasing or decreasing as a result of IBM’s HR initiatives.
Johnson Controls is a diversified technology and industrial company with 180,000 employees and customers in more than 150 countries. For this project, APQC interviewed Wendy Hirsch, executive director of workforce analytics, at Johnson Controls.
Johnson Controls has a workforce analytics center of expertise within its HR function. One of the team’s first predictive projects was to understand why voluntary turnover was slowly rising. The workforce analytics center of expertise discovered that, at Johnson Controls, missing performance management milestones such as yearly goal planning and performance assessment is predictive of voluntary turnover.
SAS is a business analytics and software provider with more than 13,000 employees and customers in 141 countries. For this project, APQC interviewed Jennifer Nenadic, manager of enterprise analytics services at SAS.
At SAS, HR uses workforce analytics to address human capital management issues and opportunities. Workforce analytics is a partnership between the HR and IT functions. One of the first predictive workforce analytics projects that SAS conducted sought opportunities to improve the organization’s already low employee turnover rate. Out of the project arose a model that predicts whether HR process changes are likely to decrease turnover risk at SAS.
T a l e n t A n a l y t i c s , C o r p . uses data science to optimize employee performance and attrition for high volume, individual performer roles including Call Center Reps, Insurance Agents, Sales Reps, Sales Engineers, Bank Tellers and the like. Much of our work predicts top and bottom performers pre-
hire. Our predictive scoring algorithms are beautifully and easily deployed into the employee sourcing, recruiting, hiring and operations processes via our award winning cloud platform, AdvisorTM.