Big data is changing the way employee benefits – and advice – are delivered through the workplace, says Capita Employee Benefits chief data scientist Dr Eric Tyree. John Greenwood hears how
When Capita Employee Benefits appointed Dr Eric Tyree to the new role of chief data scientist in 2014, the organisation set off on a path towards more informed benefit programmes, a smarter approach to voice-of-the-member communications and, one day, maybe even algorithm-based financial advice.
Big data analytics has been steering retail strategy for years, but employee benefits is only just beginning to tap its potential. Speaking to Tyree, one gets the impression that data mining can offer real efficiencies for employers and trustees, and is something EBCs should have been engaging with years ago.
“Bringing modern people analytics into the benefits space means looking at employees as consumers of employee benefits. These techniques have been around for ages in retail. Family life, income and age are huge predictors of the benefits people will find attractive,” says Tyree. “Employers can save huge amounts of money by steering their benefits spend in the right direction – between £500,000 and £1m for a medium-size company.”
Tyree says employers engaging with people analytics for the first time often don’t even know their workforce. “Typically the data is an eye-opener as to the kind of people who work in an organisation. This gives the hard numbers and the ammunition for the human resources director to go to the CFO to say they want to change the benefits,” he says.
One obvious example is group life insurance, where it can be worth ditching the product for younger workers, even if the scheme becomes more expensive as a result.
“If you are a high-street food retailer, most staff are not interested in life insurance. So you are better off paying more per head for the policy for the 20 per cent of people who want it, or offer it on a voluntary basis,” says Tyree.
Another quick win can be on flex, he says. “There is no reason why you cannot customise the flex window to change the order of what is offered, based on your data on individuals. So if you are young, maybe the discount vouchers or the iPhone 6 offer is at the top, and if you are older, it will be pension,” he explains.
Capita EB has built modelling based on 15 years of academic research around engagement with pensions. It divides a typical workforce into three groups. At one extreme are the disengaged, passive savers, often with low incomes, who will always take the default. At the other extreme are the highly engaged people who need quality self-select options.
“In the middle are the ‘limited personalisers’, typically comprising around 40 to 50 per cent of the workforce. They will engage, and only need a bit of a nudge to do so. So they might increase their savings, but they are also likely to move into self-select. If you are an employer, these are the people you want to be looking at,” says Tyree.
Tyree sees a growing role for data analytics in filling the advice gap. “The Amazon ‘people like you’ approach is perfectly achievable in the employee benefits context and is tied to robo-advice, which is about to explode,” he says.
“The current set of robo-advice is a process designed by pension consultants, and it looks like it. But in the future it will be looked at in a big data way. This will involve ‘people like you’ recommendations that will all be big data driven,” he says, though he remains silent on the issue of whether Capita is preparing to launch such a proposition.
When it is pointed out to Tyree that ‘people like you’ recommendations are technically advice, he says: “This is all doable. The solution is simple. It is like training a robot to be human. If I train a human, I maybe monitor them for 50 per cent of the time at first, and reduce that as their expertise improves. You can do that with a robot. And if the regulator loosens the definition of advice, it will be even easier.”
