HR must ensure any data it uses takes contextual information into account, according to a roundtable discussion hosted by Capita and HR magazine.
The roundtable on data-driven HR brought together nine HR professionals to discuss use of data and analytics in the function. One of the key themes was the need to make sure HR understands the relevance and context of people data, and understands which stats are important and which are not.
Graham Salisbury, director of people and organisational development at the Tony Blair Institute for Global Change, likened the situation to motorbike riding. "If you're doing a high speed motor race, like the TT, you might have the information about your speed to hand, but that doesn't matter," he said. "You only need to know your rev count. You might assume you need a certain set of data in your business, but you find actually that you do not."
Penny Newman, Lewis Silkin's chief people officer, agreed that data can be misleading. "Imagine supermarket delivery drivers, where their employer might keep a track of how long it takes them to make deliveries," she said. "You might assume that the faster driver is going to be more efficient, but actually it could be the one who is taking more time and interacting with the customers more who is adding that value and doing a better job."
Contextual information can help HR to correctly interpret people data, added Salisbury. "I was working at an aerospace firm once, where turnover was around 4%," he said. "I ran a project where the turnover was 2%. So is that good? In fact, in context, it was pretty bad. A charity might have a turnover rate of 30%, but when they compare themselves to others in their sector, they might realise that's actually not as bad as it first appeared."
Alex Tullett, head of benefits strategy at Capita Employee Benefits, said that information from external sources can help to add context. "We found that Middlesborough had the highest turnover, but when we looked at information from Rightmove, that put it into context for us," he said. "You could see it wasn't a problem exclusive to one business [but due to the location having a high turnover rate generally]."
"People can often add two and two, and get nineteen," Helen Giles, executive director of people and governance for homeless charity St Mungo's, warned. "You might, for example, notice from your data that your BAME (black, asian and ethnic minority) colleagues are more likely to be given a disciplinary, and then you might assume that means your organisation is institutionally racist. While that might still be the case, there could be many other reasons behind it that this piece of data isn't showing you."
A full round-up of this event will be published in November's edition of HR magazine
