The topics of BigData in HR is white hot. This week we launched our BigData in HR research study, which describes precisely how world-class organizations build high performing talent analytics functions. This research represents the culmination of many years of work, talking with hundreds of companies and many vendors in the market.
In this article I’d like to highlight some of the findings, and invite you to join our research membership program to get more information. You can also come to a free webinar highlighting the findings on June 26 at 2:00 PM EST/11AM PST.
Five Basic Findings:
Let me summarize the research and the maturity model in five basic findings.
1. Most HR organizations have many standalone analytics and measurement programs, but have yet to pull it all together.
More than 80% of the organizations we talk with have lots of great measurement programs (learning measurement, a talent acquisition analytics team, and often a workforce planning team), but they are rarely pulled together. This means that what most companies measure is the effectiveness and efficiency of HR itself.
While this is important, it pales in comparison with the potential to correlate all this HR and people data to the actual running of the business. What sources of hire predictably deliver the best performing sales people? What management factors directly cause high levels of theft in retail stores? What training programs or skills directly correlate with customer retention and upgrade?
Companies have been answering these types of questions about marketing for years. Most advanced marketing organizations know precisely what impact every marketing program has, and how it generates leads and revenue and market awareness.
HR, by contrast, is still often focused on measuring itself.
This is not to say that many companies don’t do an excellent job of measuring employee engagement, for example, and using that data to find management gaps and organizational problems. But ultimately when all this is pulled together, into what we call a “level 4 predictive analytics” team, the business results can be transformational.
2. Despite the existence of many HR software tools and systems, tools are not the answer.
I spent many years selling software, and it’s very easy to get people excited about the next great tool set which delivers talent analytics. Both Oracle and SAP are starting to heavily market their offerings, and virtually every other talent management software vendor and most tools companies are now selling tools targeted toward HR.
What our research shows, however, is that tools are not the answer. No company ever has all its people-related data captured in one place (so don’t plan on it happening). What world-class companies do is develop their own “data dictionary” and put in place a strategy to collect, aggregate, and use data from many sources. They standardize on tools, working with IT, and they learn to use them well. It is far more important to standardize on a set of tools which is supported by IT than it is to constantly search for the “next best thing.”
I’m not saying that tools are not important – they are. But tools are not the solution. Many companies have undergone multiple failed projects to build HR data warehouses or other similar programs (I was just meeting with one such company yesterday) and what they find is that the answer is a focus on process, strategy, and patience. Which leads me to point #3.
3. There is a predictable maturity model to world-class analytics.
As we detail in the research, there is a 3-5 year maturity model to building a world-class analytics solution. And this is a journey, not a destination. Just as Wal-Mart’s incredible retail analytics systems were not build in a short period of time, similarly the people-analytics solution must be built piece by piece over time.
Our research shows four major stages to success, and these are highlighted below:
Briefly, what the model shows is that organizations must move from “reactive” to “proactive” and then from “strategic” to “predictive.” As the research shows, you can try to “skip” steps, but it typically backfires. You can turn analytics into a “project,” but then you have the problem of inconsistent data, lack of reliability over time, and ultimately a lot of extra effort.
On the other hand, if you build a function over time, the talent analytics “function” will become highly strategic and actionable. The goal here is to turn analytics from a “project” to a “process” and from a “team” to a “function.”
4. Don’t try to boil the ocean. Stay very business focused.
There is a natural tendency to try to build analytics from the data outward. While this type of architectural approach is necessary (and you have to get the data dictionary in place), the most important thing to do early is find a single business problem to focus on.
In Levels 1 and 2 you are building infrastructure to move from being “reactive” to being “proactive.” But as soon as you have a team and start to assemble tools, focus on a few significant business problems. If you focus on one or two current, relevant business problems (ie. theft in the financial services case, sales productivity in the insurance company, customer retention in the third organization), then you have the green light from the organization to really focus on business solutions.
Remember that in levels 3 and 4 you are not trying to make HR better, you are trying to use the information you have about people to make the company better. And this is best done by focusing on a well-known, financially relevant problem.
5. It’s all about the team.
The final little tidbit I’d like to share is that talent analytics solutions must be built by highly skilled, focused, passionate people. World-class analytics teams are made up of people who have deep skills in database, statistics, business analysis, and executive communication. A typical HR staff person or OD specialist may or may not fit here. Our research shows that these teams can be very small, but must be made up of the right people.
Over time the HR analytics function can take over and govern analytics in all areas (in most of our BigData group the HR analytics team took over learning analytics and other areas, but not necessarily recruiting), and they interact with other analytics groups in the company. At level 3 and 4 the team should be partnering with finance and marketing, so that you can share tools, data, and process experience.
As we discuss in our “Agile Model of HR” research, transparent access to talent information is a hallmark of a high-performing company. I think this area, BigData in HR, is one of the most exciting new career and business opportunities in our industry.
We look forward to helping you.