Predictive Analytics as a Strategic HR Solution

Sue Lam's picture

This interview is the second in a series on workforce predictive analytics. APQC interviewed John Callery, Director of People Analytics at AOL. In this role, he focuses on building a predictive analytics team to ensure talent-related decisions at AOL are driven by data and science. In this article, he provides insights into the field of predictive analytics as it relates to talent and makes recommendations on how HR can leverage analytics.

APQC: Could you tell me about your background and how you got into predictive analytics?

Callery: My educational background is in applied mathematics and business analytics. I started my career as a systems engineer with an aerospace company at NASA’s Langley Research Center. After a couple of exciting years there, I began a career in consulting. My roles in the consulting world got me interested in HR, specifically around how my background with data and applied mathematics could apply to helping people, and making sure companies are better supporting their employees and building strong organizations. That started me down the path towards workforce analytics. My role at AOL is focused on taking them from a reporting focused HR team to being a more forward looking, predictive, and strategic partner for the business when it comes to anything that has to do with our people.

APQC: How do you define predictive analytics to an HR professional?

Callery: To me, predictive analytics is using scientific concepts to find patterns and make predictions that can help make traditional HR processes more relevant for employees, and make the HR team in general a more strategic partner with business leadership.

APQC: Why do you think it’s important for today’s HR functions to leverage predictive analytics?

Callery: Everywhere I’ve ever worked there has been an “arms race” for talent. People always want to have the best, the brightest, the most talented people. [Companies want to] find them, recruit them, and keep them as long as they possibly can. [They want to] keep them happy and productive. The importance of talent has always been there, and the importance of predictive analytics is to support people in a better way. Right now, a lot of this is being done without any eyes on the data and without an eye on personalizing experiences for employees. What once required very broad strokes based on best practices in the industry can now be tailored a lot more. For example, even if a company has a low attrition rate, they don’t just want to say, “we expect the attrition rate to be X over the course of the year.” Ideally, teams want to target individual businesses and job functions, and know not only the [attrition] rate, but what motivates each person in order to institute more targeted practices that will improve retention for all employees.

APQC: What kinds of projects or questions can predictive analytics help HR answer?

Callery: I think predictive analytics should be embedded into all HR projects and all questions that HR asks, whether it is predicting who is going to be the most successful candidate for a job, figuring out what benefits to offer, or what changes to make to an organization. All of these should be based on predictive analytics or at least have a strong foundation in the relevant data. We work on a number of projects across these areas and I’m fortunate to work in an organization that wants to have as much data as possible to make the employee experience better. We are using [predictive analytics] specifically on the career side and on the recruiting and retention side to figure out how we bring in the best talent, and how we recognize who might be at risk [for attrition] and intervene to ensure our people know how valued they are at AOL.

APQC: What do end-users and HR professionals need to know to leverage predictive analytics?    

Callery: I think the basics are usually around transparency and communication. Predictive analytics shouldn’t be radically changing the way you interact with employees. It should help you do the things that you are already doing or should have been doing in a better way. A lot of times, it’s really just giving the HR and business leaders information, coaching them around what that information means, and helping them with the communication and action plans. You should also be asking what your analysts know about your HR professionals and businesses as much as what the HR professionals know about the analytics. If an analytics team can be partnered with the end-users in the businesses and talk to them about what they’re hearing day to day and what results that they are seeing, it allows them to be more informed and do better analyses having that contextual information.

APQC: What characteristics do good predictive analytics functions and analysts/data scientists have?

Callery: From an analytical group standpoint, there are several things you need to have. You have to have a group with good mathematical and statistical science-based backgrounds who know how to do the hard work of manipulating the data, finding the insights, and understanding what’s actually going on. You need to have some level of programming and computer science skills in order to manipulate the data and implement models effectively. You have to meet at least this bar to get into the game. What I look for within my team is also a few other things. The biggest piece is knowing how to ask good questions. There are all kinds of things that can be looked at from an analytical standpoint but they may not be the best questions to ask if you’re trying to drive change in an organization. Knowing how to ask good questions and how to communicate with the business in order to find out what’s happening and to provide information to support action are critical things that I’m looking for.

One example around hiring is that someone might ask an analytics team to focus on describing what top performers in a company look like so that they can hire more people like them. An analytics team needs to be able to refine that question. Instead of simply defining top performers, the team should be looking at how and why they are different from the rest of the organization. This can lead to many insights, not just for hiring, but also for changing the organization to help current and future employees reach top performer-level potential. These are very different questions with very different results.

APQC: Do you see any common misconceptions when working with HR and predictive analytics?

Callery: There are a couple of misconceptions or fears about predictive analytics in HR. From a misconception standpoint, people think that predictive analytics gives you an answer. It doesn’t necessarily do that. It helps you to better understand the world and the system that you’re working in, and it might help you find better solutions than you had before, but it’s not some perfect answer that’s going to be one solution that fits every situation. We try to talk about that a lot with our teams and make sure they understand that what we are doing is trying to facilitate the decision-making that’s happening, not provide a blind solution. There are reasons why you might not want to strategically act on a single piece of information. For example, you might have data that says, “here is what your top talent looks like.” Instead of taking that and trying to build a homogeneous organization that looks exactly like those employees, you might instead want to build out a more well-rounded team that looks very different from this group in order to match current and future needs, and ensure that your team can stay successful long into the future. Hopefully that’s what [HR] is doing. It is not only about understanding what the data says. It’s about understanding what the goals are too.

APQC: What are three tips that you could give to HR functions that are starting up their predictive analytics work?

Callery: Whenever you’re starting a predictive analytics function, a lot of people default to diving into the data. [People want to] start figuring out what fun and cool nuggets they can pull out to show people. I think that’s the opposite of what you should be doing. The first thing that I would recommend when starting a predictive analytics group is to go out and talk to the business.  Talk about what they are hearing, what their challenges are, what their strategy is so you can know what areas to prioritize when you start looking at the data. What do you focus on? What are the high impact areas that can make a big difference? Having that understanding and engagement with the business is something that I’ve found builds much better partnerships and allows you to do much more with the resources that you have than you would otherwise. Step one is to go out and talk to the business leaders and people in the organization to make sure you are connected to them.

The second piece is to take the information you’ve gathered from people and pick a couple of pilot projects. Select a couple of projects where you can focus on one or two questions that can make a significant business impact with teams that you believe will be champions for you. If you’re doing HR analytics work in a silo without the support from the business, you’re not being successful because the whole goal is to improve the business. You’re not going to be able to grow and make an impact in the organization if you don’t have champions within the business leadership. Find a couple of problems, identify the champions for whom the problems will be solved, and partner with them to solve the problems and communicate with them about the problem and get people excited about it.

The third piece is about diving into the data. Make sure you have clean data and make sure you understand the constraints that you are working under. Those are all important things that I think are more traditionally found in an analytics group but they can be more easily prioritized by having better conversations earlier and strategically selecting the first questions that you focus on.

APQC: What predictions do you have for the future of predictive workforce analytics? What capabilities might become available ten years from now?

Callery: I think predictive analytics will have a big impact on HR. Similar to the impact that it’s had on finance, marketing, and pretty much every other area of businesses, I think having analytics being built into the HR function is going to be very important to their future growth and the strategic vision. A lot of HR leaders talk about having a seat at the table and being able to give strategic advice to executives. Predictive analytics is going to be to making that happen.

Ten years down the road, I think that there are a lot of exciting opportunities in this space. I think that one of the things that you’re going to see hopefully sooner rather than later is that more HR analytics teams will be partnering with finance, strategic planning teams, and all areas of the business because everyone needs better information about the people within an organization. As HR builds its predictive analytics functions, partnering outside of HR is going to become even more critical. By working closely with other teams and building out a broader set of data and business understanding, HR will become more recognized as a leader for all people-related topics inside of a company. Instead of seen as a separate support function, it will become embedded in all strategic decisions and become a critical driver of productivity and success in the marketplace.

You can hear John speak about how AOL is using predictive analytics as a strategic HR solution at the Predictive Analytics World for Workforce conference in San Francisco on March 31st. PAW Workforce is offering friends of APQC a 10% discount. To register, click here and enter APQCWF15 to take advantage of this great offer. 

 

1 Comment

Anonymous's picture
Interesting contents and motivations!