Big Data & Predictive Analytics in Talent Acquisition


In this video, Dion Love, a Principal in CEB’s Advisory Group, shares key insights and best practices around how HR organizations can better leverage big data and talent analytics to make better business decisions.

Watch this on-demand webinar now to learn:

  • How to improve the relevance and actionability of talent analytics data to drive competitive advantage
  • A step-by-step, scalable approach to strategically investing in and improving talent analytics
  • How to transform HR’s impact on business metrics through proactively addressing key business challenges

Webinar Transcript:

Scot: Joining us now is Dion Love, Senior Director, Product Strategist at CEB HR.

Dion: Thanks a lot Scot. Good morning, good afternoon, good evening everyone wherever you are. Welcome to CEB’s presentation on Talent Analytics for Talent Insights, “Big Data and Predictive Analytics in Talent Acquisition.” I’m Dion Love, and I’m the Principal in the Advisory Group at CEB. I’m delighted to have the chance to share with you an overview of key insights and best practices from our combined work across CEB on the practice and power of better talent analytics. In the 21st century, organizations collect a huge amount of data, and that data is capable of providing insight to inform and drive better business decisions. Data promises better insight into our economics, our customers, our product and more. Talent is no exception, but today few organizations are able to translate that data into real insights than business leaders’ trust and act upon. In fact only 41% of today’s leaders tell us they are confident that talent data can help them to make the kinds of decisions that truly impact their business in meaningful ways.

While most organizations try to boost the impact of data through investments in technical capability like better data platforms or more analytic talent, the best organizations focus alternatively on improving the relevance and actionability of talent insights with a clear goal of helping the business make better more informed business decisions. They invest first and foremost in the capabilities and tools that help them to do that. Let’s go ahead and make a start over on the next page where I’ll show you the highlights of that broader sea they research that brought us to this conversation today on talent analytics. Talent data availability or the volume of data on and around talent has exploded across the past 20 years, and top lot, that data availability is placing significant pressure on organizations to turn that raw data into meaningful insights to inform talent decisions. As you see now on the bottom of the page, not surprisingly, CEOs are placing greater expectations on talent insight to inform their business needs to be competitive today.

What we’ll see around the course of our brief conversation here today is how assessment and related data can help you and your organization not only to meet that expectation, but to drive competitive advantage through talent advantage. Let’s go ahead and dive in over on the next page. I think what Socrates used to say, “The beginning of knowledge is the definition of terms.” With that in mind, we are going to start with a couple of quick definitions around two terms that are fundamental to what we are going to be speaking about here today. These are on the page in front of us here. At a high level, we think of two concepts as instructive and helping us to understand the analytics landscape where we are and where we want to go. You see them arranged here on the page in front of us, descriptive analytics and prescriptive analytics. Descriptive analytics on the top of the page tell us how the world is today and we use them to inform decisions leaders might choose to make about where they want to go.

An example here might be based on the current state of engagement in market X. We might want to prioritize that for a new product launch of product “B” for example. So it’s data-driven, setting the leader or leaders up to decide. You see descriptive analytics in a number of forms, dashboards, benchmarks, analysis and forecast. Predictive analytics or prescriptive analytics you see down the bottom, you see pushes a bit beyond that. In prescriptive analytics, we apply more advanced statistical techniques like modeling to understand where the world is likely to be tomorrow and importantly, to prescribe specific actions that leaders should take, as you see appearing in different forms as well, predictive models, driver analyses and more. Each of these types of analytics has its place of course, but to make the best possible of the business decisions, we want to have as many errors in our quiver as we can. Descriptive analytics and prescriptive analytics both can help you to forecast and plan for the future.

For most organizations, the frontier is first building an approach to better talent analytics overall. Let me show you a couple of steps our CEB research has shown us are going to be key to achieving that. This is over on the next page. Organizations’ investments need to be increasingly targeted to not just generating more reports on talent, but in providing better talent analytics. This page shows you what we think that migration looks like. It used to be on the left hand-side of the page that when we talked about HR metrics, we were talking mainly about what you see down the left hand side here. We were focused on primarily providing information, providing it on-demand or at-request basis, and that usually resulted in a set of reports or a set of talent metrics that were reported to leaders across the organization. That’s what we call talent reporting.

Compare that though, to talent analytics. That’s what you see on the right, where the purpose is not to provide information, but to improve business decisions. We are not just responding reactively to requests for data, but we are proactively trying to address key business challenges, and the outcome we are focused on is not a metrics report, but informed and demonstrable impact on the business. Why would we want to do that? Some compelling reasons are over on the next page. Organizations that apply talent analytics move faster and act more decisively. Surprisingly as you see on the page here, it says top priority in HR for almost half of organizations today. Why? Further on the page here you see the most competitive [inaudible 00:06:45] realize that access to data isn’t a problem solver in and on itself, the insight that comes from that data is where competitive advantage truly lies.

For those that do this well, the benefits are real. Take a look here over on the next page. Just two examples here that widely reported in The Business Press, here the headlines only, first across on the left hand-side, PNC through a more thorough approach to higher source analysis identifies a superior talent source in internal hires, and accordingly realigned its hiring guidelines and saves millions of dollars in improved sales through getting better talent into its critical organization segments. Second across on the right hand-side, DHL assesses graduate candidates for key skills using an assessment that gives them deeper insights faster than their talent competitors. How do we get there? You must be surprised. Just getting more sophisticated is not going to be our best answer. Take a look over on the next page.

Focus on analytic sophistication alone will get you increased outcomes, but a singular strategy focused on analytics sophistication will provide diminishing returns after a certain point. That’s what the graph on the page here is showing us. Let’s take a look a little closer detail. That certain point is pretty much exactly where the average organization is today. Let me orientate this graph then we’ll talk a little more about the conclusion we can take from it. First, the horizontal axis represents the continuum of analytic sophistication from low to high. By analytic sophistication as you would see in a definition around that term, we mean an organization’s effectiveness in using advanced technology and methodology for talent analytics. If you look to the vertical axis, there we are looking at the impact on talent outcome. So there we see talent outcomes, like for instance engagement, performance, retention and more, but once you pass that point of what I’d call mid-maturity, you’ve gone from low to good, you don’t see a lot of returns in simply focusing on going from good to great in analytic sophistication alone.

There’s still something missing, that’s going to enable you to get to ultimately higher returns in terms of improved talent outcomes. That’s what we wanted to understand through our research at CEB. How do we get higher returns’ more impact from talent analytics? What’s the rest of the answer? Let me show you that over on the next page. There are three keys to getting to the right answer here. Overall, our goal is to transform HR’s impact on the business through better talent analytics. In doing this, we need to address three key challenges that organizations face today. Those challenges are running across the top of the page here, criticality, capability and credibility. First on the far left here, we need to address a criticality challenge. Most companies tend to first prioritize a range of metrics that they then try to apply to addressing their business challenges, or they find themselves simply reacting to ad hoc requests for data regardless of its criticality across the business.

The best companies focus first on prioritizing the critical business questions that need to be answered, and then they determine what HR data and analytics can best answer the most questions. That leads to both higher business impact and more scalable application of our analytic resources. So that first challenge is around identifying and focusing on the most critical analytics for your organization. Moving to the middle of the page, we also need to address the challenge around capability. By that we mean HR staff capability in analytics and how a sole or even primary focus on quantitative skills we’ve seen is insufficient, that the capability gap for HR staff is not around quantitative skills alone, but it’s actually more around business judgment. So for a given level of quantitative skills, for a given level of analytic capability where we as an organization need to focus staff capability development first is around the application of business judgment to that pre-existing level of quantitative skill.

So our imperative here is to help our HR staff to apply business judgment to data signs. That’s what we wanted to understand again in this part of the initiative. How do we build the right HR staff capability to get the impact that we know and we see the best companies achieving from talent analytics. The last challenge is across on the right hand-side of the page there. Last, but not least, we need to overcome a credibility challenge. That is a perceived lack of credibility of our data of our insight coming from our analytics in the eyes of our business leaders. We do that by not only building on our capability in the most critical analytics, but by driving end user ownership talent analytics. Having the confidence to be able to put forward not only a solution, but to allow business leader to test that solution, to ask questions around that solution and to have his or her HR counterpart respond to and help him or her to look at different scenarios across the set of solutions that HR is providing.

Each of these of course is a conversation onto itself. A broader research addresses each one of them in turn, but for now, let me turn us into the talent acquisition context that we are focused on today to see how through CEB and CEB is a job talent measurement support, organizations address these three priorities in improving analytic impact. This starts over on the next page. I’ll take us straight into the detail here. Starting with some background data that comes directly from CEB SHL Talent Measurement’s work on predictors of job performance. You see here on a scale on the left hand-side of the page of zero to one, we see the different predictors of how high a performer, an individual hire will be in their new role. You see starting down the bottom at zero, we have random prediction. At the top where we score one, perfect prediction. You see each of the dimensions, each of the approaches that we can take to understand potential job performance of a new hiring role.

Things starting from phrenology on the bottom of the list, very close to zero there, moving up through educational qualifications, employment interviews, group exercises or on the job type simulations, right up to assessment centers at the top. Obviously, seeing this data you can understand why a core part of CEB SHL’s focus is around the delivery of effective and impactful assessment methodologies. So that’s where we first start to focus in this part of the discussion. What I’d like to do here is turn over to the next page to show you where this takes us, understanding that these types of approaches, assessments, group exercises, structured interviews are on the top end, at the top of the range in terms of their predictability or their predictive power in understanding new hire potential job performance. Here’s where we see the benefit coming from assessments in particular.

Take a look across the page here. We are showing how quality of hire established through, for example assessment centers, gives us not only a reliable understanding of that individual employee’s potential, but it’s our fastest way to profitability and savings. Take a look over on the left hand-side where we are comparing two organizations here, organization A and organization B. Organization A and B each hire approximately 2,500 employees each year, and the average salary is 75,000, average cost of benefits, around 25% of salary. Organization A selects well, organization B, not so well. You see the impact of talent outcomes from that excellent verses poor selection in the boxes on the left hand-side of the page here. Organization A has a turnover rate of just 10% and a performance rating of 80%. Organization B on the other hand has a higher turnover rate, almost 50% higher at 14% and a performance rating much lower and only 61%.

See how this not only though impacts the performance of the individual and role and therefore the performance of the organization, but it also impacts the bottom line. Take a look over on the right hand-side to see the annual cost differentiation between organizations A and B. The additional performance penalty that we see coming to organization B, as a result of that poor selection, you see there on the screen in front of you, a compelling reason to focus on data selection based on greater predictability around performance using assessment centers. Let’s take a look over on the next page though to see the other piece to solving this problem. Not only do we see opportunity to improve the way in which we select and assess talent, and we’ve seen a more reliable way of doing that through focusing on quality of higher defined by potential to perform in role which we better understand through assessment centers and assessment methodologies, we also see that the recruiting function itself has some room to improve in terms of positively inflecting the quality of hires coming into our organizations today.

See on the left hand-side of the page here, we see the overall argument, concerns on applicant quality and recruiter efficiency results in concerns with turnover and underperformance. On the right hand-side here, you see some of the top level data that comes from our research at CEB for specifically, the CEB recruiting Leadership Counsel Membership. You see concerns here specifically around, firstly on the top left, 72% of applicants considered low to average quality by the recruiter. Only one in four hiring managers report that recruiting actually influenced their selection decision, leveraged their labor market expertise, their understanding of broader talent strategy or the part line building capability to get them a better hire through, not just responding to the hiring manager’s demands, but influencing them towards the better talent selection decision.

And then top right-hand corner, 65% of hiring managers are not satisfied with recruiters’ impact on their business, that’s not surprisingly when you consider the preceding data here on the page. Also as you look to the bottom of the page, some broader top level results that these produces, new hire turnover much higher across that group of new hires compared to all employees, and then bottom right-hand corner, new hires today are underperforming. One in five new hires are considered bad or regretted decisions by hiring managers. The graphic on the page here kind of under-serves the true impact of that poor hiring decision in today’s new work environment, more metrics, more networked environment where new hire performance, or in fact any employee’s performance not only impacts the tasks with which they themselves perform, but also impacts the performance of those around them.

That one bad hire is having a much greater impact on the organization today than you would typically have seen in the past. Take a look at the bottom of the page though, this is where the true impact really hits home in terms of the impact on recruiting budget. New hire turnover results in more than $1.6 million in rework for every 1,000 hires. So some significant room to improve there both in the way we assess talent and in the way our recruiting functions are run in order to not only conduct a better assessment, but to leverage talent analytics in getting better, higher quality hires into our organization. Here on the last page of the presentation is an overview of how CEB SHL Talent Measurement and the broader CEB organization supports you in re-inventing talent acquisition.

Three key priorities that we see here are essential to not only understanding talent analytics, not only moving from descriptive analytics to prescriptive analytics and finding the best mix of the two in order to best address today’s business challenges and specifically, talent implications of those challenges, but also to ensure that we are getting better quality talent into the organization to drive today’s business results. You see across the page here those three priorities, transforming recruiter capabilities, building smart and more strategic recruiters, not just focusing on building quantitative skills, but helping them to apply business judgment and influence hiring managers. In the middle of the page here, leveraging predictive analytics, so enabling strategic workforce planning to be done especially where it matters most, leveraging new insights from big data.

Our CEB TalentNeuron group helps you to understand current trends and opportunities across the global labor market and develop acquisition strategies to drive quality of hire at the very top of organizations’ talent funnels. Last, and across on the very right hand-side, once we have that new hirency [SP], how do we most effectively ensure that we are driving his or her performance to the optimal possible level? A range of solutions there across, helping you to measure what matters, assessing your talent not only for today’s performance, but how well you are setting them up to be a top performer tomorrow, understanding and measuring for potential. Not just skills, their potential to rise to and succeed in more senior critical positions across the organization.

Last, but certainly not least, how to support and build an on-boarding program that sows the seeds of high performance throughout their tenure with your organization is much more than a checklist of activities they need to undertake in order to orient themselves into the final position they have come into with your organization. So that’s the overview of talent analytics as we’ve understood across our broader research in the CEB Organization and also how we bring that insight together to help you to understand and leverage the true potential of talent analytics in your organization across transforming recruiter capabilities, leveraging predictive analytics and driving new higher performance. Let me finish up by taking us to the last slide here where you’ll see my contact details. If you have any questions, comments or feedback or if there is anything you’d like to say to continue with today’s conversation around better talent analytics, please feel free to tweet @dionlove #TalentInsights, and we’ll get right back to you with any questions you’ve got.

We’ll also be happy to continue the conversation there, but at this point, let me go ahead and end my formal comments. I wanted to thank you all for your participation in this slide session today. At CEB, we know what the best companies do. We help you to do it too.