Data and Wellbeing


How analytics could unlock health gains for your staff

Much-pilloried retailer Sports Direct is not particularly renowned for its employee wellbeing practices. But in August, it emerged that it had been making some efforts in the area. Employees who began shifts in its warehouse were asked to press a button featuring either a happy or sad emoji; if they pressed sad, they would be asked whether they were sure about their decision. Press it again, and they would be called to see a manager to discuss their reasons for doing so.

Unfortunately, this didn’t amount to an outpouring of concern from a business condemned for the ‘gulag’ conditions on its sites. While the survey was supposedly anonymous, staff were being identified by their fingerprints, prompting howls of protest from unions and a fresh barrage of woeful publicity.

Sports Direct may have been misguided, many would argue, but in seeking to use data to shape its approach to employee wellbeing, it is treading an increasingly well-worn path. A survey by Aon Employee Benefits last year found that 72 per cent of employers now use some form of data to drive their health and wellbeing strategy.

This isn’t surprising in itself. After all, we are enjoying a well-documented explosion in employee data, along with a growing recognition that investment in wellbeing pays off in both the short and long term.

While the most popular sources of data are well-established – absence statistics and employee engagement surveys – more and more companies are looking at fresh inputs, such as data from Fitbits and other wearables, calls to employee advice lines and even ‘sentiment analysis’ of internal social media channels. “Done responsibly, this is a huge opportunity for HR to understand the impact the organisation has on its people, and how to best support them and create an environment where there’s a balance between their health and their commitment to work,” says David D’Souza, head of engagement and London at the CIPD.

There are several factors driving the trend, according to Mark Witte, principal at Aon Employee Benefits. “There are cost pressures on employers and the impact that poor health could have; support from the state is not what it used to be so the health burden has shifted; and we’re now presented with a greater array of choice than before in terms of private healthcare, specialist providers and different pathways we can offer,” he says.

Proponents of wellbeing data analysis argue that flagging up potential health problems across the worker population could prevent serious issues (and costly insurance claims) down the line, too.

BNP Paribas, for example, recently carried out a workplace stress and wellness study of more than 500 of its employees, who wore a Microsoft Band 2 that measured various biometric measurements such as heart rate, as well as how many steps they took. The data was gathered and analysed by health technology start-up BioBeats, deploying metrics such as whether employees’ perception of stress tied in to their actual stress levels and whether interventions such as breathing exercises made a difference.

It found that stress levels were 23 per cent lower where employees followed breathing exercises on an app, and ‘high rumination’ (where people consistently worry about work outside of work) was reduced. “Traditionally, HR and occupational health have relied on subjective data from questionnaires, so there’s a certain bias. We bring in objective physiological data – the amount of sensors we track builds a picture of wellbeing that would be impossible through questionnaires,” says David Plans, CEO of BioBeats.

“Many employers have pathways already such as an EAP or GP referral, but in an ideal world we’d respond before this stage. There can be stigma around admitting you’re not coping and, by the time some [mental health issues] present themselves, it’s at a late stage.” Using data to predict workplace wellbeing trends can avoid costly interventions down the line, he argues, and even reduce the risk of chronic diseases.

All of the employees involved in the wellbeing study chose to take part and knew what their data would be used for. “It all goes back to intent. Employees don’t want oppressive oversight,” says D’Souza. “If it’s being done just to improve productivity, people may feel uncomfortable not opting into something – the relationship becomes unhealthy.”

A recent survey by consulting firm PwC found that 65 per cent of people feel technology has a role to play in their health and wellbeing, yet 38 per cent do not trust their employer to use the data it collects to benefit them.

Employers need to be upfront about how the data will be collected and used, advises Jocelyn Paulley, director at law firm Gowling WLG. “Would an employee be comfortable using an employer-sponsored health provider if they thought news of their condition would get back to their employer, which may make promotion or work assignment decisions with this information in mind?” she asks. Next year will see new data protection responsibilities under the General Data Protection Regulation (GDPR), which places more emphasis on employees’ consent to data being used – so it’s an area where employers need to tread carefully.

Gaining insights from aggregate or averaged data, rather than zoning in on individuals’ activities, is the safer path. “We would only ever give an analysis of data across a large population,” says Peter Blencowe, managing director at Bluecrest Health Screening. “Some companies may ask whether we can slice it by department, but say you have 10 people and one has a certain condition – that’s going to be pretty identifiable.”

Mapping data from sources such as EAPs, insurance claims and wearables also has inherent limitations because it will never cover the entire workforce, adds Ivan Robertson, co-founder of business psychology firm Robertson Cooper. “With any voluntary collection of data, you run the risk of getting a biased picture. EAP data will only show you who’s picked up the phone – what about the other 80 or 90 per cent of the workforce?”

Initiatives where staff sign up to wear a fitness tracker or take part in a data collection exercise don’t give a full picture, either. “You’re only looking at a slice of the workforce that was already interested,” says Gemma Milford, a health insurance consultant at IHC. “More data might not change their behaviour, but having a more inclusive overall wellbeing strategy – for example, offering fitness classes or encouraging people to take the stairs – might have a greater, albeit anecdotal, impact.”

Professor Sarah-Jane Cullinane, assistant professor of HRM and organisational behaviour at Trinity College Dublin, believes that making employees aware that you’re monitoring them could end up having the opposite effect. “They could worry that it’s just another way they have to be a ‘good employee’,” she says. “It becomes another pressure, and there are so many outside variables that make it difficult to judge fairly.”

She adds that regular ‘pulse’ questionnaires, taken anonymously, can provide a more accurate picture of the happiness (if not necessarily health) of the workforce. “If it’s anonymous, people are more likely to be more honest. If they know they’re being watched they’re more likely to change their behaviour. Coming up with an aggregate can show trends without people feeling they’re being watched.”

For employees, the question of ‘what’s in it for me’ will always prevail. Law firm Shakespeare Martineau wanted to make its staff more active as a tie-up with Commonwealth Games England and offered employees Fitbits. It set a challenge for people to collectively track enough steps to get to Australia’s Gold Coast and back again – the equivalent of 20,000 miles. Two participants will win the chance to see the 2018 Commonwealth Games in Queensland.

“We are honest and upfront about the data we collect from the Fitbits, and while the Gold Coast goal is there, 90 per cent have said they will continue to use the watch once the challenge ends,” says Joanna Thornell, director of sales and service excellence. Many of the benefits have been anecdotal, she adds. “Teams are hosting walking meetings and finding them to be more productive. It’s not just an excuse to clock more steps for your team.”

And while small behavioural changes are gratifying, collecting data on wellbeing can also offer hard evidence on the efficacy of existing health programmes and suggest where businesses’ health spend could be better targeted. “Lots of companies are trying to define their wellbeing strategy, not just because they want to ensure it’s effective, but to make sure they’re doing it for the right reasons,” says Jo Salter, director in PwC’s people and organisation business.

Beth Robotham, head of business development at Bupa, says data is critical in demonstrating return on investment in wellbeing: “When it comes to creating a sustainable wellbeing strategy, it has to deliver value to both the employer and the employee. Data provides a clear sense of what is important to their people and helps them decide where to invest to make a meaningful impact,” she says.

The golden rule is that data usage has to be both well-defined and adequately explained. “Any wellbeing statistics can inform the wider HR strategy in conjunction with other data, but pushing for more and more data could create ethical dilemmas,” says Neil Mountford, chair of the UK Employee Assistance Professionals Association.

You only have to look at examples in the US where employees have been microchipped so they can buy food and drink or log into computers, or a start-up offering predictive analytics so employers can decide what proportion of its staff are likely to become pregnant, to see there is the potential to go too far.

There is also the question of how greater employer knowledge of health risks in their workforce will affect insurance premiums. More and more corporate health insurance providers offer data analysis as an add-on, “so clients understand where their money is going and they feel they’re getting more value from their spend”, says Witte.

At present, insurers can only look at data at an aggregate level and do not use predictive analytics around individual workforces to calculate premiums. The GDPR sets clear boundaries on how individual employee data is stored and used. But it’s equally clear that in the longer term, businesses could come under collective pressure from insurers (as they have in the US) both to pass on more detailed information about their workforce and to undertake specific interventions to keep their premiums down.

In the meantime, making tactical decisions about groups of at-risk employees and running programmes that can help with their medical or physical health could prevent problems – and therefore claims. For employers that stay on the right side of the line between paternalism and Big Brother, wellbeing data remains a world of possibility, even if the future is less clear.

Published by People Management, 27 Sep 2017