Big data can help us make sense of absence; Managers should develop ways to predict when, why and ideally where staff are most engaged

July 15, 2013 4:19 pm

Big data can help us make sense of absence

By Andrew Hill

Managers should develop ways to predict when, why and ideally where staff are most engaged

It’s hard to hide from big data. After submitting its staff records for independent analysis, one retailer discovered it was paying more than 150 employees who had called in sick years earlier – and simply disappeared from the workplace. The human resources managers responsible were so terrified of the potential backlash if they revealed the figures to their chief executive, they decided to keep the job of handling absences in-house.They should have gone ahead and outsourced the task. The long-term benefits of knowing where your staff are – and more important, why they aren’t where they should be – far outweigh any short-term pain. The benefits should also override employees’ natural suspicions that the company is bringing in Big Brother to vet their sick-notes.

Absence management companies have been around since the 1990s. But they have grown from a small group of consultants, trading anecdote and hearsay, into analytical hubs aggregating information to prove or disprove HR’s suppositions. After a sunny weekend of barbecues and parties, for instance, you might assume unscheduled absences from the office would increase. Data for the UK generally do show an upward spike in “sickies” on Mondays, according to David Hope, chief executive of FirstCare, a London-based absence management company. But small heatwaves don’t trigger more unauthorised days off than normal, because the impact of cold and flu viruses diminishes during the summer.

Other revelations can be specific to individual staff – such as the man who claimed leave to grieve for a dog that call-centre records showed he had already buried the previous year – or more general. In Scotland, an apparently random series of peaks in staff absence was explained only when they were correlated with the professional football fixture list: Scottish sick days, it turned out, coincided with the aftermath of matches between arch-rivals.

Absence management companies operate call centres staffed by nurses and other experts on behalf of large employers. Employees call the centre rather than their team leader, explain why they’re off work and receive health advice. Their manager is automatically notified.

It’s all too easy to dismiss this as a system for snooping on potential malingerers. If it’s seen that way, it “creates all sorts of toxic effects and [the system] becomes useless because people are gaming it”, says Julian Thompson of the RSA, a think-tank. He is co-author of a new report with Vodafone that predicts UK organisations could benefit from significant cost savings and output gains by instituting better and more flexible working practices.

Reducing unjustified absence is obviously one aim. But properly used, such a service can be far more valuable. It removes the risk that soft-touch team leaders will grant extra days off, or that martinets will force their underlings back to the production line when still ill. It ensures proper records are kept – useful in case of dispute or dismissal. It also nudges companies to improve management practice. When the staff member comes back to work, the call-centre prompts the manager to conduct a formal interview. This is the sort of workplace conversation that too often is dropped in favour of a passing remark – “Better? Let’s get on, then” – but can in fact uncover deeper-seated physical or mental health issues.

The data can be equally powerful. For example, FirstCare spotted that bus drivers with shifts starting midway through the day complained of back problems more often than those who drove the first shift. By extending the handover time between drivers so the incoming crew could adjust their seats properly, the bus operator alleviated the problem.

In the right hands, data should be a tool to make workers’ presence more productive. Clever managers should now develop ways to predict not just when and why employees disengage, but also when, why and ideally where they are most engaged with the task in hand.

HR managers often complain they are trapped by the board’s perception that they are mere payroll and personnel monkeys. Well, here’s a strategic role for them: building an empirically sound understanding of how people work best. Then it would be possible to use data not only to reduce unscheduled absences, but also to ensure that when staff are there, they really do turn up.