Just like any other function in the organisation, HR is finding itself increasingly under pressure to make use of big data and analytics. A 2015 research report from Bersin by Deloitte says that HR needs to harness what it calls ‘out-of-the-box’ predictive analytics technologies to help it make better decisions around talent. Such software could help organisations improve retention, strengthen leadership pipelines and augment employee career paths. Rather than relying on sophisticated data scientists to crunch the numbers and interpret the statistical findings, the report urges HR leaders to take charge and do it for themselves.
The problem is that when it comes to adopting new technologies, HR has traditionally been regarded as a laggard. But there are pockets within the wider function that are deemed more progressive than others, namely reward, or ‘comp and bens’. As reward intersects with a number of other key disciplines, including talent management, employee engagement and learning and development (L&D), it is well placed to demonstrate how HR data can be used strategically.
“Using the data available can ratify the level of engagement and strength of the employee proposition,” says Philip Hollingdale, founder and CEO of Staffcare. “This will have a beneficial impact on a number of areas including informing and influencing the recruitment and retention policy, and ensuring that the reward strategy is hitting the targets set by the board.”
In principle, reward has a better chance than some other areas of HR of winning the appreciation of the board when it comes to using big data. As Adam Nuckley, reward consultant at Innecto Reward Consulting, points out, reward already speaks in a language that is readily translatable to the business because everything it talks about has a value in pounds. He adds, however: “Ultimately, what [organisations] need to be doing is combining this with other information.
Indeed, it is the combining of multiple databases that effectively turns it into ‘big data’ and allows it to be used more strategically. Historically, however, this has been a challenge for siloed HR. Datasets have typically only been partially interrelated, which has made it difficult to effectively link or indeed model with any degree of statistical reliability.
But the emergence of more sophisticated technologies here is making this much more doable. “Technologies can now capture greater amounts of data, structure it in an accessible way and integrate it with hitherto unrelated databases to create big data,” says James Markham, managing director at employee benefits technology provider SBC Systems.
“This big data offers the possibility of applying analytic techniques to uncover stable correlations within particular populations, which will help refine reward design to better achieve its objectives.”
In some reward departments this is already beginning to happen. The University of Lincoln has implemented a new integrated HR and payroll system which has allowed it to incorporate a far more comprehensive reporting suite. “There is a passive understanding that data to inform decision-making is the correct approach and that the data becomes more powerful when it is overlaid with other information,” explains Ian Hodson, reward and benefits manager. “Our new system has fed data triggers in to populate other systems, and we have a much more cross-function approach to data collation and production than ever before.”
Hodson says one of the “exciting enhancements” of having the new system is ensuring that he and his team can draw a correlation between base pay and consistently good performance: “The aspect of the link between loyalty and base pay is an interesting one as statistically we have seen a competitive marketplace where those changing jobs have had a powerful negotiating position, often creating a significant differential against longer-serving staff.”
The university also has monitoring in place from a pensions perspective, and the lowering of the annual allowance has meant it gives far greater consideration to the timing of pay reviews. “We are also looking to have flexibility in how we reward to ensure the right outcomes for employee and employer,” says Hodson.
Neil Morrison, group HR director, UK and international companies at Penguin Random House, reports that his organisation has used a “huge amount” of data and analytics to examine how reward should be shaped and benefits structured, and to assess the balance between fixed and variable pay. This followed the merger of the two publishing companies in 2013 along with their individual reward and compensation platforms.
“That’s [involved] everything from understanding people with different backgrounds but the same job titles, to the take-up of benefits and the specific value of certain ones,” he says. “Whether they have the value you think they do and whether that has links with turnover and retention and whether therefore the investment is adding value.”
As Penguin Random House was seeking to redesign its benefit offering, it needed insight on take-up levels in general but also how it applied to different demographic groups and what part salary levels played. “We saw younger people weren’t taking up private medical care, so were able to offer the flexibility of a lower-value product,” says Morrison.
According to Staffcare’s Hollingdale, organisations typically use data analysis both before and after introducing a flexible benefits scheme. Using benchmarking and return on investment data, including potential tax and National Insurance (NI) savings, helps secure buy-in from the board. And once introduced, the success of the scheme can be accurately determined. Where an organisation has holidays and absence tracking as part of its benefits portal, realtime data and reports help it understand absence trends and whether its reward programmes are influencing engagement.
“Increasingly clients are also using our survey tools within the platform to get employee feedback on benefits,” says Hollingdale. “Not only does this give employers valuable insight, but it also helps as part of the engagement process, as employees feel like they are being listened to.”
Meanwhile, Penguin Random House is working with Innecto to examine the structure of its incentives and base pay as well as how to balance pay structures in new and old roles. Both internal and external data is proving invaluable: “Being able to take external data and analyse and compare it with our internal data and make decisions on pay structures will hopefully take us forward for the next five years rather than just working on individual pay structures,” explains Morrison.
Looking ahead, University of Lincoln also intends to use data analysis to inform its pay committee about links between senior-level performance and reward. Hodson stresses that, as with all remuneration committees, there is a focus on ensuring that budgets remain controllable and fit for purpose.
“One common theme that you often see where a non-consolidated bonus forms part of the reward package is a position where a bonus will be given over a number of years but there is no review of progression within grade banding or the sizing of the role,” he explains. “In higher education and because of the links to final salary, this can often directly impact on the longer-term growth of pension funds. We will be looking to our new system to generate monitoring reports to look at individual outcomes of performance reviews and correlate this against base salary reviews. This will ensure that the right data is put in the hands of the decision-makers.”
As Nuckley underlines, reward already talks the language of business and this should help the function begin to demonstrate the value of the information that resides in HR and payroll, as well as other databases. In turn this may help secure senior buy-in for supporting other parts of HR to invest in relevant data analysis expertise, tools and resources.
Markham reckons reward professionals still have some way to go though and fires a broadside that some have a “retail consumer” view of IT, which has led to a market focus on specific rule-driven calculations and screen “look and feel”, rather than “underlying functionality”. He warns that this has limited the ability to apply and promote the development of the algorithms and heuristics required to produce strategic insights and actions from data. “There is a need to build a stronger base of understanding to enable the industry to move forward more quickly,” he warns.