The first order of business for any HR department is keeping the company’s best performers happy, productive and growing so they don’t walk out the door. But when you consider how businesses have historically made most of their talent management decisions, you’ll find a somewhat scary mix of decisions based on guesswork, gut feel and anecdotal conversations.
Before the advent of Big Data, it was tough for employers to analyze and optimize things like their hiring patterns, employee compensation and retention. Thankfully, those days are behind us! But the process of collecting, interpreting and then acting on Big Data can still be a daunting task for many employers.
Some of the biggest challenges businesses encounter when they embark on using a data science approach to managing their talent include:
Getting a good grip on internal data
Every department and location across an organization may define its data differently, making it more difficult to interpret. Once the data’s collected from all parties, it then needs to be cleansed or normalized so it makes sense companywide. For example, if you want to see how your company’s vacation time compares to the industry average, you’ll first need to ensure every internal party is uniformly tracking vacation time separate from bereavement and sick days so you can make an accurate organization-wide assessment.
Leveraging external data
Smaller companies may not have the infrastructure to be able to gather enough data to identify patterns that matter. But they can leverage historical Big Data from surveys, research and studies, and real-time data such as weather forecasts, or partner with a third party that can offer access to its Big Data.
Having a data analyst on staff
Data analysts in the HR space have only recently become a trend. Today, HR leaders need to be able to identify the key business issues the company’s trying to solve to help decide which data points should be tracked and analyzed in the first place. They then need to be able to summarize the data and simply explain to the company’s stakeholders what the data is showing -- and why it matters.
The good news is this: There are many tools out there that are making it easier for companies to pull data together so they can start seeing patterns and spot potential issues to help make better-informed business decisions. Think about some of the common questions companies often want answered: How can we predict who will be a successful employee? What’s causing high-performing employees to leave? How often should employees be moving internally to maximize engagement and retention?
Once you have all of your own data and external data in shape, you now have the reference points to answer these questions and understand what you could be doing better. By using a data science approach to managing their workforce, companies can gain insight into the most probable outcomes so they can take action.
For example, a company experiencing higher-than-average turnover may want to know whether it’s a trend being experienced industrywide or if it’s specific to their own firm -- and why that is. The data may reveal that the people leaving have been in particular roles for twice the number of years compared to the industry average, causing them to seek new opportunities. By knowing that the chance of attrition could increase 40 percent by keeping an employee in a role twice as long as the industry benchmark, an employer can then develop retention and career development plans to keep the best talent engaged.
Once companies start using benchmarks to see how they stack up to their peer networks, they can then build predictive models that flag employees who may be at the highest risk of leaving in the future. With that knowledge, they can then either part ways with underperforming employees sooner or create a plan of action to keep prized employees on the team.
It’s important to recognize that actionable predictive analytics cannot be accomplished with data from only one company -- you need benchmarks to compare yourself to peer organizations and a massive data set. Consider when you go to Netflix® to rent a movie and recommendations appear based on your buying patterns. The site looks at all the movies rented by its entire customer base, and then identifies the people who generally like the same movies as you. You’re now getting actionable information based on choices made by people similar to you that can help guide your decision.
What’s the future of Big Data?
It’s not only having all of this data at our fingertips to guide better decision making. It’s about having the data “pushed” to us in a predictive manner when we need it the most. Think about when you’re walking to your car after work and Google® Maps tells you how long your drive home will take. You didn’t have to request the information, the app knew it would be useful to you at that point in time. Eventually, your HR solutions will do the same.
If you’re experiencing turnover, while processing the termination the system could prompt you with candidate resumes who would be a good fit. Further, if the data flags several employees in New York who may be future flight risks, an automated assistant could schedule an appointment with them on your calendar to coincide with your next trip to that office.
When you consider how critical it is for a company to retain its top talent, embracing the insight gained from data and analytics becomes a no-brainer. While it may be intimidating at first, consider how great it would be to manage your workforce to meet your specific needs, gain insight into what you should be planning for in advance, and always make sure you have the right skills available at the right time.
(About the author: Marc Rind is vice president of product development and chief data scientist of ADP Analytics and Big Data)