Predictive analytics has long been considered the apex of HR analytics. Although speculating about the future is never a bad thing (it is, in a manner of speaking, my line of work), viewing predictive analytics as the end goal may be looking at things the wrong way. The message many of these graphical representations send is that if you aren't using predictive analytics, you haven't fully realized your potential as an HR analytics operation. Conversely, if you have the ability to do predictive analytics, then you need never fear, as you have reached the nirvana level of data mining and all problems will begin melting away in a soft, backlit glow.
However, the evidence to support this theory is not there. Plenty of companies have been around long enough to have developed top-level analytics programs, but still struggle with data issues (security is at the forefront, but usage is also a big part). There are also many smaller companies that have been able to leverage the people data they possess for maximum results, despite not having the technical infrastructure to support big data or predictive analytics.
This is because predictive analytics isn't really the end-all-be-all of HR data. Predictive analytics, and this might be heresy, isn't even all that difficult--some of the models that exist are plug and play--but what is the final goal of any people data endeavor? It should be to help the business succeed, and predicting turnover is not the only way in which businesses succeed.
In fact, this highlights the glaring weakness of predictive analysis; it is a one-size-fits-one approach--tailored people analytics--very useful for answering a specific question (e.g., will the organization have enough people to achieve our goals in three years?), but quite useless for answering others (e.g., can we expand into China?).
For the majority of HR problems, predictive analytics are not the answer. Daily issues arise that are more easily solved with reporting, correlations, or some other simple analysis. The trick lies in knowing which tool or approach to select for the job (and in having access to the right tools). This is where the concept of adaptive analytics comes into play. By having the right mix of analytic talent, analytical tools, and team structure, problems that can be solved using people data are solvable quickly.
Solving problems quickly should be considered the true endgame for anyone in the people analytics space. Having complex models is necessary for some questions, but not for others. Sometimes the simplest solution is best, and if HR analytics teams are able to solve a variety of problems in a variety of ways, they are truly poised for success.
When i4cp investigated the components of the most successful HR analytics teams, four main themes were reasserted throughout, all of which are characteristics of high-performance companies (and explored in the report The Rise of Adaptive Analytics, now available exclusively to i4cp members).
- The use of adaptable analytics technologies
- A focus on integration of technologies
- Usage of analytics for leadership and risk assessments
- A separate reporting function
The first on the list, the use of adaptable analytics technology, is a good example of the adaptive analytics method. Traditional business intelligence software, such as SAP, performed well (i.e., had a positive correlation to market performance and analytic acumen), but the standouts were RapidMiner and R (R in particular was a top-rated analytical tool, yet only 5% of respondents reported use).
If you are unfamiliar with R, it is a statistical computing language that has the dual advantages of being open source and having strong global community support. The open source background means that not only is it free, it's also easy to adapt for multiple types of uses. The community support means that there are ready-made packages that have specialized capabilities, helpful for the complex problems presented by people data.
In the Adaptive Analytics study, flexibility was the one trait that all of the highest-rated analytical software had in common. Flexibility allows the user to mold the tools to whatever form is best suited for the problem at hand. This provides for as many variation on the types of solutions as there are on the types of problems that HR analytical teams must solve.
This is what adaptive analytics is all about: having the mindset and the capabilities to react and adapt to the challenges that arise, as opposed to forcing the problems to fit the skills and tools that are available. Perhaps even worse is the erroneous idea that data analysis is not the purview of HR, and thus investment into the development of HR teams is unnecessary; this is a proven path to failure and the very antithesis of the adaptive mindset.
To succeed in business, HR leaders must see that the way forward involves not building up, but building out. A variety of tools and expertise is the recipe for a highly adaptive analytics teams (as spelled out in more detail in the report The Rise of Adaptive Analytics, now available exclusively to i4cp members). The companies that succeed will be the ones that embrace agility and adaptability--at least, that's the prediction brought on by the analysis.