RSS Feed

Feature

More Articles: Latest Popular Archives

Grabbing data by the cornucopia
Print – Issue 165 | Article of the Week

Alex Cresswell
confidential

 

Each month we will be sharing four, carefully-chosen articles from the Latest Issue of our flagship publication ‘theHRDIRECTOR’ which exemplify the high standards we strive to archive. We hope you find this in-depth article of interest and decide to become one of our valued Subscribers.


“If we have data, let’s look at the data. If all we have are opinions, let’s go with mine”. So said Jim Barksdale, CEO of Netscape back in the 1990s. Today, there is certainly no lack of data, we’re drowning in it – IBM estimated last year that 2.5 quintillion bytes of data are produced every day – the big question is, what actually is it we want to accomplish?

Article by Alex Cresswell, Managing Director – Pymetrics

Nobody can doubt the rising significance of data! This year, the annual HRcoreLAB event in Barcelona added an additional track to its agenda focused specifically on HR analytics. The auditorium was packed for every single session – as clear a sign as any that the topic of data is firmly-established within the HR departments of European businesses. It is tempting to think that this has come about suddenly, so perhaps it is helpful to remember that ten years ago, the topic we were all focused on was business intelligence. Five years ago, BI was considered passé and so everyone began talking about big data, and now it’s predictive analytics and AI. It is clear that whilst AI has been in development for decades – the term was first coined by eminent computer scientist John McCarthy in 1956 – we have only just reached a point where the available technology, combined with our understanding of what that technology can deliver for us, is beginning to affect the pace of change, and that pace is catching some organisations by surprise.

“Instead of first identifying the problem and then determining which assets will enable that problem to be solved, the starting point seems to be to look at the available data, before deciding what to do with it. This is not how to get the best value from analytics”

Listening to some of the commentary today around artificial intelligence, including that from well-respected figures such as; Bill Gates and Elon Musk, you could be forgiven for thinking that the sky is about to fall in: Mass unemployment leading inevitably to the point where machines decide they can run the world more effectively without humans. Desperately trying to avoid further hyperbole, it is safe to say that we are living through a data-driven epoch, where technological change is have a huge impact on our work and home lives. But we can try to gain some perspective on this by looking back at similar periods of technological and social upheaval. During the late 1700s, for example, the Industrial Revolution led to riots in Britain when the large-scale automation of our manufacturing industries caused many jobs to become obsolete. However, that wasn’t the end of the story; new roles were created as a result of industrialisation, people found new opportunities and incomes went up. Today we are also forecasting elimination of roles; PwC has suggested that 30 percent of roles in the UK will be automated out of existence by 2030. We can expect that society will cope with change this time around in a similar fashion, as evidenced by the rate at which new roles are being created – the jobs that some graduates are moving into this year did not exist when they began their degree courses, the likes of – community manager, social media manager and content manager. It is, therefore, more helpful to look past the headlines and to think more pragmatically about how data and AI can support and enhance our businesses.

Suddenly it has become imperative for companies to be able to show that they are “doing great things with data and analytics”! However, in too many cases, the prevalence of buzzwords does not disguise the lack of clear thinking or business value. Instead of first identifying the problem and then determining which assets will enable that problem to be solved, the starting point seems to be to look at the available data, before deciding what to do with it. This is not how to get the best value from analytics – we need to begin by defining the important questions – then we can work out how the data can help. Conversely, living in a world filled with data presents us with some complex ethical and moral questions, as highlighted by the recent Cambridge Analytica scandal – as those well-worn words of wisdom stated; “with great power comes responsibility”, and abuses of power often cause catalysts for big change. Indeed, up to now, people have generally been willing to provide data in order to access apps or services they are interested in, without considering how that data may one day be utilised. And, for the people who are manipulating the data, as they become increasingly abstracted from its source, it can be difficult to remember that the individual who originally provided that data may not like what it is being used for. We can be sure that the organisations buying data from Cambridge Analytica were not thinking about whether the people completing Aleksandr Kogan’s personality test on Facebook would approve of what they were doing. As controls and measures go, however, the new General Data Protection Regulation (GDPR) is unlikely to change this, as people will still need to provide data to access the services they need and will struggle to imagine the myriad ways in which that data could potentially be used in the future.

That said, GDPR will help by enabling people to delete their data if they become aware that it is being used in a way that they don’t approve of –  but by that time, of course, it may be too late – the customer is unhappy, trust has been lost and the potential fall-out includes significant reputational damage to the organisation. In a business and HR context, of course, damage – reputational and otherwise – is a critical concern. We are using information provided by employees who did not supply their data because they wished to access a particular service, but because they had to, as a condition of their employment, or in order to get their work done. The perceived balance of trade is already unequal. This means that we need to take even greater care to maintain trust and to use data appropriately. For example, it is now possible to determine which employees are thinking about quitting by looking at their email data, their LinkedIn updates or their attendance patterns. But once staff realise that this is happening, there is a collapse of trust and business performance is impacted. We can learn some lessons by looking at how some progressive companies think about data. For instance, LinkedIn has a set of values that employees operate by. The first of these is ‘members come first’ and this is kept in mind whenever the company thinks about novel ways to use member data.  This is because LinkedIn understands that without the trust of its members, it essentially has no product or service to offer. Another test that LinkedIn regularly uses, although this one does not have the status of official company ‘value’, is to check whether a proposed use of data avoids the ‘creepy’ test. It would be helpful to adopt similar approaches, when thinking about using the data of your own employees.

Many companies, of course, are already highly sophisticated in their use of data and analytics; they are driving business value and doing so in a way that contributes to the engagement of their workforce, without ever making them feel uncomfortable.  Indeed, one particular major industrial company has put together a team to answer strategic questions such as; “what’s the impact of variable pay on performance”? They are accomplishing this by looking at the data they already have in their systems and complementing it with secondary data, such as; surveys and assessments. Philips is looking at data from the employment market to determine where they will locate their facilities based on talent availability. Meanwhile, the well-established analytics team at Unilever have a healthy approach, where they will use checks and balances on what they are seeing from the data, to ensure that inaccurate information does not end up taking them in an unhelpful direction. Interestingly, another business identified which roles in the organisation were most similar to others, so that internal mobility and workforce planning could be more effectively managed. Using the data, they were able to highlight previously unrecognised issues within a management community that enabled development efforts to be more precisely targeted. It is clear then, that there are interesting and highly-beneficial use cases out there for data and analytics and today’s modern HR function absolutely needs to build the skills to understand how to work with the data available to it – or face being left behind. The key to whether this will lead to competitive advantage – or just become a huge distraction – is how well you decide which questions to ask.

pymetrics.com 


If you found this article of interest and would like to receive our flagship publication each and every month (in print or digital), please Subscribe

Receive more HR related news and content with our monthly Enewsletter (Ebrief)