Sergey Nivens - Fotolia
Human resources analytics is the shiny new toy for many companies looking to procure and secure talented workers. But before jumping on the trend, companies need to understand and plan for what analytics can do. Paul Seo, global head of HR business intelligence and analytics for Marsh & McLennan Cos. Inc., recently discussed HR data analytics with SearchFinancialApplications.
What are the best practices for planning and evaluating an HR data analytics project?
Paul Seo: To enable analytics, the first thing you need to have in place is infrastructure. It gives you scale and speed. The amount of data you are handling is exponentially growing -- the speed to provide an answer is [faster and faster]. Infrastructure gives you a highway system to move a lot of data, synthesize it and push it through much quicker to meet the demand.
Second is data governance. There is data coming from everywhere, and it is also going out everywhere. You need a traffic cop in the middle, a governance process in place to make sure people using analytics are using it as intended. Otherwise, it is a highway system without police or traffic lights.
The third is your analytic capability. You need the right talent in your HR organization. This is what creates the business value. Turnover went from 10% to 15% last year; now what? You need to be able to provide insight that will allow people to make the best decision possible. Without a proper skill set in-house to turn data into intuitive and compelling stories, it would be difficult to make HR analytics stick.
The fourth is cultural readiness. With enough money, you can build the infrastructure [and] hire enough smart scientists to build the analytic capability, but if you don't use it, it doesn't really matter. To truly gain knowledge and the value of analytics, you must integrate it into the decision-making process.
It is a very different mind-set to make decisions using analytics versus simply enabling analytics. It is not something you can switch on and off instantaneously. It is something that is embedded.
Unless there is a cultural readiness and cultural capability to translate and leverage the analytics, then the first three don't really matter. I would argue that this is the biggest impediment to making HR analytics work. Applying analytics to our workforce is a new concept in HR, and -- naturally -- this brings wariness, and hence we are our biggest obstacle to embracing and utilizing analytics. Until we get over this fear, it will be extremely challenging to truly apply and utilize analytics in HR.
Do you have any advice on identifying metrics that offer the quickest payoff for HR data analytics?
Seo: It is the integration with finance that holds the key to the biggest payoff: monetization of HR functions. Organizations should know exactly how much they spend to hire. They should know exactly how much they lose as a result of termination. As they invest in talent development, organizations should also know the expected return. [There is a] cost to hire, cost of replacement and cost of turnover. If you are going to provide training, for example, what is the average return you are going to get on that training? We need to put dollar signs next to everything we do.
The HR culture is grounded in relationships. How do you change that DNA? How can companies get over their reluctance to apply HR data analytics and live with the results?
Seo: It has to be mandated by the CEO. Human capital management is a business. We deal with the business of people. To put it very simply, running a business is about maximizing your profit. So we should approach human capital management in the same fashion. All HR functions, including talent acquisition, talent management -- I like to call this workforce optimization—and talent retention should be driven with optimization in mind, and optimization, to put it very simply, is about maximizing your profit.
For every single person who walks out the door, for example, we should be able to tell the business what the bottom line impact would be and what we are willing to spend to either retain or recruit. Yet most HR organizations don't follow this recipe.
Why are HR leaders hesitant to apply analytics?
Seo: I am not sure hesitance is the right word. I think the fundamental ways by which HR functions are evaluated are changing, and that naturally brings uncertainty.
When you quantify everything, you will be evaluated based on not only the improvement itself, but also the degree of improvement. This makes everything very factual; so, less about the relationship and more about the results. That is daunting for many in HR leadership. It presents a lot of political pressure they never felt before, and I think many HR leaders are simply trying to figure out how to navigate this new expectation.
In Deloitte's Human Capital Trends survey for 2015, only 8% of the respondents surveyed believe their organizations have a strong HR data analytics team in place, a slightly higher percentage than in 2014. Why is that?
Seo: Most HR leadership is trying to figure out what to do and how to navigate this new paradigm shift. The problem isn't technology or the analytics. I think it is purely cultural.
Top-down is the only way to make it stick. A chief HR officer needs vision and needs to see analytics as an addition, not a subtraction.
What software does Marsh & McLennan use for analytics?
Seo: We use Oracle OBIEE. We went live about a year ago. It is a visualization tool. It takes the data in the data warehouse, spins it and puts it out in a presentation in a dashboard to allow end users to consume.
Everything is on premise. Our data warehouse is on prem and OBIEE is on prem. We don't use cloud -- not yet.
We use PeopleSoft for core HR functions.
What are the pluses and minuses of OBIEE?
Seo: The pluses are that it is backed by a very powerful engine. It also uses the same language as the transactional system, like PeopleSoft, which is a very widely used application, and Taleo, a talent acquisition application.
A minus right now is visualization, in terms of aesthetics. It is not quite there yet with some other competitors, but I think Oracle is making good traction.
Do you have any lessons learned regarding your analytics software implementation?
Seo: Change management is a lot more complicated than building a dashboard or an analytical capability. Many organizations underestimate the difficulty of changing habits. It takes time to marinate.
What are some examples of the possibilities of HR analytics?
Seo: Predictive analytics is important for controlling turnover. It provides the trends of certain individuals who leave, their attributes, where they come from, their length of service, their location and job code. You can factor all that into the existing population and try to understand who is most likely to leave in a certain period of time. Based on that, you can reduce the likelihood of them leaving. There are some top actions HR can perform to reduce turnover.
Another would be optimizing talent acquisition. Analytics helps you understand historically where you have obtained the best talent. If you need engineers, you will know where you have had the best luck. Based on when the vacancy is likely to come, you can recruit and build a pipeline. It will reduce your cost of hire and your vacancy rate.
Get more advice on implementing HR analytics
Learn about trends in predictive analytics
Read a handbook on HR analytics