Can HR add value to its data and analytics?

When it comes to HR, is data and analytics just another passing fad? Although scepticism abounds around measurement, its ability to create useful insight should not be underestimated.

In 2008, Google took another step towards omniscience. The company claimed it could forecast outbreaks of flu, weeks before the US Centers for Disease Control and Prevention, by intelligently using search data. For a while, it succeeded. Google had authored yet another big data success story. But, earlier this year, researchers from Northeastern and Harvard universities uncovered a rather large problem with Google Flu Trends: it was no longer accurately predicting flu cases – overestimating the number of cases in the US for 100 of the previous 108 weeks, by up to twice as many.

And the reason for this inaccurate reporting? Google’s own search algorithm and its auto-suggest feature. Type ‘Do I have f’ into Google and it suggests ‘Do I have flu’, which may have led to people searching for flu when in fact they were interested in knowing if they had food poisoning or even fibromyalgia (a chronic condition that causes pain all over the body).

Google’s focus on correlation rather than causation (scientists are clear that “correlation does not imply causation”) led to what the researchers termed an embarrassing case of “big data hubris”. In other words, it forgot some of the core principles of statistics and mixed big and small data in a way that proved problematic.

Can business and, in particular, HR learn anything from the parable of Google Flu Trends? Of course, most HR departments are nowhere near as sophisticated in their use of data, but there’s no ignoring the steady march of HR analytics. The question is: is anyone actually getting it right? Or is the profession in its own small way following Google’s precarious path, seduced by the possibilities of oh-so-sexy big data and attempting to measure too much, collecting data on meaningless things or starting its explorations from the wrong place?

Time to talk about analytics

According to Laurence Collins, director of HR and workforce analytics at Deloitte, it’s time to reconsider HR’s relationship with data. “I’ve spent the past few months at conferences taking the temperature of HR around analytics,” he says. “It’s been hot for the past couple of years, but I’m not seeing any significant progress in the profession as a whole in terms of actually being able to do anything about it.”

Many HR directors would agree with this. “The measurement of the impact on people is the Holy Grail of HR,” believes Jeremy Campbell, HR director at HR and payroll specialist Ceridian. “I’ve read a lot of books and articles about the methodology. What I’ve not been able to find is the practical application. There’s a lot of hype. We get slightly seduced by the potential of what the data could tell you as opposed to the reality.”

“This is a really complex area and the minute you start talking about big data and HR analytics, everyone wants to get involved and there’s not much clarity around what you actually mean,” adds Jez Langhorn, SVP and chief people officer at McDonald’s.

Given HR’s fondness for fads, it’s easy to be cynical about breathless HR analytics chatter, often coming from the technology vendors with the most to gain. “Data is the new commercial,” sighs Neil Morrison, group HR director at Penguin Random House. “When HR had to be commercial, it was all about understanding this, that and the other. Now, it’s all about understanding data. But it’s not just data that’s going to help you. The gap for me is around intelligent use of data.”

Why data matters

It’s a confusing subject, and a sensitive one too. Why does it actually matter? “There’s a massive risk for HR [if it doesn’t engage with data],” says Edward Houghton, research adviser for human capital metrics at the CIPD, who is leading the Valuing Your Talent initiative (see p9 for more on the project). “We live in an environment of data everywhere. Running a business without understanding data is not how the 21st century works.”

Collins makes a similar point: “There is a culture of analytic rigour being applied to everything, from smart devices in your washing machine to the way we work. Structured and unstructured data is there for us to use. HR either embraces that or decides it wants to be replaced by finance. If HR doesn’t seize this opportunity to become more credible around measurement and the insight that comes from that, we will be replaced by functions that will.”

For Campbell, the purpose of his search for data that proves HR’s impact is to help him come up with “a clear articulation of the business case”. He says: “HR professionals do a fantastic job of helping people reach their potential, but I’m not sure how good a job we do of articulating the business impact of that.”

“If you really value people, how do you demonstrate that?” adds Houghton. “It’s about having a conversation. Sometimes metrics will allow you to have those conversations with those leaders. If you don’t have the measures behind you, you can’t have that argument.”

At PwC, for example, HR has worked with finance to “create advocates in the finance teams for HR reporting,” says Chris Weeks, strategic workforce planning leader. “If finance colleagues see the step change in the HR reporting, then they point people to HR for HR data, rather than provide it themselves.”

Paul Kearns, co-founder of the Institute for HR Maturity, who also chairs the British Standards Institution’s committee on professional HR standards, says his interest in metrics was formed when he worked as an in-house HRD at a manufacturing company. “What really bothered me as an executive HR guy was that I couldn’t convince my exec colleagues that HR was important,” he recalls. “To convince them, I first needed to convince myself by measuring what I did.”

Then there’s the move towards integrated reporting (IR), which HR magazine has covered in-depth and increasingly looks like being a matter of not if, but when, and the ongoing discussion over HR professional standards. IR requires reporting around people-related measures, particularly human, intellectual, social and relationship capital. It makes sense for HR to play a leading role in reporting on these – whether or not it will be able to do so is another matter.

Meaningless measures

Getting all this right is easier said than done. Kearns puts it bluntly: “Most HR people are not measuring anything meaningful. There’s this big game being played where HR measures its activity and tries to use that data to justify its existence. The problem is, it’s not impressing anyone.”

We all know the adages ‘You can’t manage what you don’t measure’ and ‘What gets measured gets done’. Maybe it’s worth coming up with a new one: pointless measurement leads to pointless data. “HR is under so much pressure to measure that people measure everything that moves and hope they hit the target,” believes Kearns. “People are running around like headless chickens and not producing anything of real value.”

Morrison expresses a similar sentiment. “There’s something in using quantitative information to gain insights, but everyone rushes out to do everything and produces a huge amount of information that isn’t very useful,” he says. “We stopped doing a lot of data reporting because it was meaningless. We were creating reports that no one read.”

Weeks agrees: “The temptation can be to develop a swathe of reports based on information that is easily accessible, but this sort of information can be of limited value and the volume of it can overwhelm and turn off business users.”

Campbell adds that he feels a lot of the numbers flying around are “HR for HR’s purpose as opposed to business purpose”. He says: “It is useful to run HR more efficiently by getting a headcount report or being able to do reward quicker, but that’s just making an HR team more efficient. That’s all well and good, but the real nugget that drives me forward is trying to figure out how to measure the impact of people on the business metrics. HR is about supporting the organisation to be better, and the purpose of the organisation is inevitably to hit business metrics.”

The simple fact is true value-add measures are harder to arrive at, which makes taking the easy route all too tempting. “It’s like that old joke about the drunk man only searching for his keys under the streetlight because the light is better there,” says Thomas Davenport, distinguished professor in management and information technology at Babson College and author of Big Data at Work. “People measure what is easy to measure.”

So, in HR terms, that means looking at cost per hire, hours spent on a training course or turnover and absenteeism. They all have their uses, but that doesn’t mean they offer much value. “These are just ‘same as’ measures,” says Collins. “You have the data, so you report on it. But if you stopped producing those metrics, what would the result be? Why measure cost per hire? What about cost per effective hire? Would you pay 5% more for someone who was going to be a good performer and stay longer, or would you rather take a chance on someone who might stay for three months?”

David Bowes, chief people officer at managed services provider WDS, is trying to use data to come up with “a more balanced view of how effective a recruitment function can be”, looking beyond time and cost to quality of hire. “I look at how successful people are in the business and in the role they were hired for,” he explains. “That’s a measure I want the recruitment people to take on board. If we are hiring people poorly, if people aren’t staying or they are not successful, that is a recruitment issue. I want to hold recruitment people accountable. That’s a more sustainable model.”

Does HR have the skills?

Tellingly, Bowes adds that he has had to change his HR team in his determination to make more intelligent use of HR data. And at every conference I’ve attended on HR analytics this year (and there have been a lot), most of the speakers have moved from other, more data-driven functions into HR in order to help HR harness the power of analytics. Many speakers have also alluded to the ‘pushback’ and reluctance they have had to deal with from existing HR professionals.

According to Deloitte’s 2014 Human Capital Trends survey, 86% of companies reported no analytics capabilities in HR and 67% rated themselves ‘weak’ at using HR data to predict workforce performance and improvement. “Very few people went into HR to work with numbers,” says Davenport, but he adds this is “not unique to HR; it’s true of marketing too” – and the marketing department is often held up as the beacon of analytics good practice.

Collins believes there are two main issues HR needs to tackle. One is “endemic” in the profession, the belief that people are not a number. “We have senior HR professionals who have been schooled in the world of talking about intangibles, the value of people and the value in individual capabilities without looking at it through a commercial lens,” he says. “The second is we haven’t attracted the more numerate and statistically minded graduates and young professionals.”

But should HR become net importers of these skills from outside the profession? It could be an unsustainable approach. “You can’t remove the function of HR from the algorithm completely,” acknowledges Collins. “To operate analytics without that level of understanding and context creates a risk. That said, there needs to be an injection of analytic capability, and maybe that means bringing people in with those skills and developing their understanding of HR over time.”

Of course, tarring all HR professionals with the ‘bad at numbers’ brush is reductive, and implying that an understanding of numbers is all they need is false. “There are HR people who are really good at numbers, and there are those who aren’t,” says Stevan Rolls, UK head of HR at Deloitte, who adds that his background in occupational psychology gives him a great statistical grounding: “It’s good to have that as statistical people always assume HR won’t know what they’re talking about.”

And whether or not an HR person is good at numbers, Rolls points out “the kind of value you get from analytics requires specialist help”, adding: “Analytics can be really helpful, but the stats behind it are often so complex, you have to be a statistician to work it. From an HR point of view, it’s having the people around who know enough to question the data.”

Morrison points out that data is “only as good as the person who uses it”. “And that’s why we’ve got to be careful,” he adds. “If we’re serious about this, we need some really different skillsets, not just to be more commercially aware. Marketing has consumer insight teams doing this work properly.”

Big insight, not big data

The real value comes not just from data, but from insight, and knowing enough not to fall into traps like Google Flu Trend’s misstep. “If you rely overly on qualitative or quantitative insight, you won’t get to the best of either,” says Rolls. “Some things you only get to by speaking to people: it’s hidden away in the relationships.” But he concedes that being able to achieve meaningful insight is “a scarce skillset”. “It’s trying to get HR out of producing MI and more about delivering insights. Too often HR is asked to manipulate spreadsheets. You need commercial and numerate people, but you don’t go into HR to be a spreadsheet jockey. The focus needs to be on the understanding.”

McDonald’s Langhorn, whose HR department is successfully experimenting with analytics, agrees. “People talk about big data, but for me it’s about big insights that come from data, and putting the information in the hands of the right people at the right time. You often get report after report, when what I want is a report that tells me what I need to know and the areas to look at further.”

Kearns believes the concept of big data has actually made HR measurement worse as people are overcomplicating things, and Collins agrees the “analytics game for HR is about trying to get much more from less”. It’s insight that matters, and that insight comes from much more than filling in a spreadsheet or implementing the latest piece of technology. It comes from understanding narrative, context and matching qualitative and quantitative information from multiple data sets. It comes from understanding the meaning of correlation versus causation, and really knowing the culture and history of an organisation. After all, as Langhorn puts it: “There is a head and a heart element to people. There’s science and there’s art.” And one cannot work without the other.