The subject of analytics is often discussed but rarely executed well, even in the most well-established talent management functions. John Sullivan and Master Burnett report
The subject of analytics is often discussed but rarely executed well, even in the most well-established talent management functions. John Sullivan and Master Burnett report
While most organisations collect data, manipulate it, and produce reports from it, few could honestly state they are leveraging HR data to drive business performance by making better decisions about how the organisation’s largest variable expense is managed.
Instead of powering decision science, the majority of established talent management analytics do a great job at telling us what has happened and how much administration has been completed. When reported, the verbal response to such analytics is often “that’s interesting,” but little to no action is triggered.
A lot of factors have contributed to the sad current state of analytics in the human resource function; luckily, most are remediable today.
Drivers of weak analytics
Skill deficiency. Until recently, many of the tools that enabled robust analytics required intermediate to expert-level understanding of statistics, a skill rarely held by the average HR professional. As a result, talent management functions lacking the budget to hire a dedicated analyst often went without. While many organisations could still benefit greatly from hiring a statistician or true business analyst, many of the tools used to power advanced analytics have become significantly easier to use in recent years.
Lack of business knowledge. The second most pervasive barrier to developing world-class analytics is one the HR profession gets chastised for on a regular basis. While more and more HR professionals are learning more about their organisation’s business than ever before, a majority lack a deep understanding of the business to figure out all of the talent management system touch-points and interactions capable of demonstrating a causal tie to business performance. The increased visibility of talent management issues in recent years is helping to rectify this, as many organisations are moving more professionals with line-management experience into human resource management roles.
Expensive tools. While a lot can be done with modern spreadsheet applications, true business intelligence platforms have been relatively expensive up until recently. The growth in availability of enterprise-class open-source applications has exerted significant cost pressure on commercial business intelligence providers, making such applications much more affordable.
Lack of quality data. Even with deep business knowledge, analytical skills in place, and access to leading-edge tools, a majority of organisations struggle to enable world-class analytics in large part because the quality of the data they have to work with is beyond poor. Because many HR professionals are overly burdened with task-oriented work, capturing data in a valid, consistent way is not a work habit that has been well honed. While workloads are not likely to decrease anytime soon, a growing number of tools in the “text analytics” market are enabling organisations to analyse trends using data in a form most talent management organisations have in abundance – unstructured text.
Complicated nature of talent management. It is easy for those outside the HR function to criticise HR, but ask any functional executive who has transferred into HR from another function and they will confirm that managing the systems that govern talent is inherently more complex than originally perceived. In many respects, the HR function is a full-fledged business within the enterprise, one that requires in-depth access to cross-functional expertise to execute well. A growing number of organisations are starting to realise that employment opportunities truly are products. As such, they must be designed, packaged, marketed, sold, supported, and serviced. This new mindset is bringing new professionals well-versed in such skills to the HR function following the realisation that it is a lot easier for other professionals to learn HR than it is for HR professionals to learn marketing, sales, or finance.
As talent management increases in perceived importance among members of the executive committee, the pressures on HR leaders to be more scientific will only become more prevalent. Luckily, several emerging tools are coming online that will help HR professionals accomplish significantly more.
Emerging tool/approach: text analytics
The first promising tool, just emerging from infancy stage and still relatively complex in nature, is enterprise class software to enable “text analytics”.
The HR function is full of data residing in blocks of text that historically were difficult to derive trend data from, including verbatim responses in learning development plans, performance evaluations, applicant/candidate/employee/manager surveys, online forums, emails, etc.
To make use of such data in the past required that human readers comb through all of the related documents, manually extracting comments in context and categorising them according to some predetermined categorisation schema, an activity capable of consuming an enormous amount of labour in even the most basic of scenarios.
Text analytics software is making such analysis available to the masses by automating the contextual comment parsing and categorisation processes. The software can complete overnight analysis that would take teams of humans manually executing the process months to complete.
The power of this emerging capability within HR is simply astounding. For instance, employment brand managers could point such software systems to blogs, social networks, online publications, and even internal employee forums to monitor for trends.
Learning and development function leaders could deploy the system across a bevy of common internal documents – including development plans, performance appraisals, performance improvement plans, business development plans, and even documented corporate strategies – to determine emerging knowledge, skill, and ability needs.
Recruiting functions could use the technology to analyse applicant surveys, new hire surveys, and online forums to determine emerging trends impacting on the performance of the recruiting process. On a grander scheme, recruiting functions could leverage text analytics to comb through thousands of academic research abstracts, identifying common hypothesis, solutions, and sources to rev up hiring initiatives.
Organisations interested in using text analytics systems must understand upfront that these systems were designed primarily for organisations to analyse consumer feedback. If you believe that employment opportunities are a product, leveraging such technologies shouldn’t be that big of a stretch.
These technologies are still very new, most coming to market only within the last year. Implementing them will be complex, just as figuring out new methodologies to use them will be innovative. In time, such systems will become easier to use. Emerging providers include Clarabridge, Attensity, Business Objects, and SAS.
Emerging tool/approach: predictive modelling
Predictive modelling isn’t new, but its application in HR is still relatively rare, and for that reason we still consider it an emerging approach. Several of the barriers that inhibit the development and widespread adoption of world-class analytics are solved by predictive modelling systems.
It is one thing to hypothesise a connection between employee absenteeism and production-line performance, then set out to prove it, but it is an entirely different thing to sit back and let statistics tell you what is related without forming any hypotheses. Manually hypothesising correlations and setting out to prove them is time-consuming and requires significant knowledge of HR, the business, and statistics.
Predictive modelling, on the other hand, requires only that you have access to massive volumes of HR and business performance data, business intelligence software, and the skills needed to set the software to task.
Predictive modelling is used extensively by marketing professionals to determine what impacts on consumer behaviour. Because predictive analytics have been around for some time, many executives already have some exposure to them and trust in their validity.
The potential applications of predictive modelling – as with text analytics – are huge in HR. Imagine being able to forecast the exact budget needed to power all HR systems at the optimal level required to produce specific predetermined levels of workforce productivity.
Predictive modelling can identify what HR systems drive increases in productivity, inhibit productivity, present risk, etc. It can be used to power decision models or simply alert you/your managers to potential issues.
Unlike the text analytics market, a majority of the business intelligence systems supporting predictive modelling are mature. The market includes common desktop software (Microsoft Excel), open-source enterprise-class applications (Pentaho), and commercial business intelligence leaders (Business Objects, etc.).
Key principles with emerging tools/approaches
Any time an individual or organisation can leverage a new tool or approach that significantly impacts on their capacity to perform, an opportunity arises to develop a competitive advantage.
Unfortunately, leveraging new and emerging tools often requires that users step out of their comfort zone. This practice is not uncommon in HR – if it were we wouldn’t be using IVR in HR call centres, workflow management systems in recruiting, or text messages to communicate with candidates.
Continuing to compete and thrive in an emerging global economy powered by evolutionary technologies that are changing at an ever-increasing pace will require HR practitioners to become significantly more strategic.
The technology exists today, as does the need to deploy it, but too many in leadership roles lack the courage to depart from the status quo.
If you look at what frustrates the executive committee the most about the current state of HR, nearly every concern could be solved with a more robust analytics infrastructure and a talent management workforce capable of understanding the business and being able to prove their impact on its performance.
Start the conversations now, give the tools a shot, start asking questions, prove what works and what doesn’t, and most importantly, have fun for a change.
Dr John Sullivan is professor and head of the HR program atSan FranciscoStateUniversity, and is a noted author, speaker and advisor to corporations around the globe. Master Burnett is the managing director of Dr John Sullivan & Associates.
