Crunching the data

As a new wave of employee communication monitoring technology kicks off, 2015 might be the year when HR analytics finally comes of age

The Australian Crime Commission (ACC) is famous for its range of high-tech crime fighting activities. They are at the forefront of data analytics to combat organised crime and its involvement with virtual currencies, illegal drug production and trafficking, and tax avoidance.

The same data analytics techniques are now being used to improve operations in another area of the ACC: organisational performance. In addition to the range of surveys that organisations typically run (internal and external), the ACC has taken the next step – analysing data that workers generate in day-to-day operations.

HR analytics is now being bolstered by a niche but growing wave of technology which sits on a company’s server or in the cloud and tracks employee communication – and in some instances employee sentiment – with pre-set metrics. For players in this market, such as Revelian and McGrathNicol, this is a new era in HR analytics.
 
“People are generating vast volumes of data in everything they do at work, but all that data is sitting there untapped,” says Peter O’Hanlon, chief marketing officer at Revelian. “What we’ve learned from other domains like marketing and logistics is that by tapping into the data that’s being generated, you can learn a lot.”
 
Indeed, many organisations know more about their customers than their employees because they’re investing in big data and are leveraging those data assets to drive their business strategy. This imbalance is now being addressed.
 
The RCAP
The Revelian Communications Analysis Platform (RCAP) looks at communication patterns (but not the specific content of emails) to offer insights into employees’ productivity, tone, and engagement levels.
 
O’Hanlon says that it’s not uncommon for organisations today to already have scanners looking for inappropriate keyword content being sent throughout the workplace – the RCAP steers clear of that. “We don’t look at content at all,” he says. “We’re just looking at the patterns of communication. We’re connected to the data sources and we then apply a range of algorithms and analytics to that high level information, which simplifies and consolidates it down to useful measures of the way people interact in the workplace.”

What can the RCAP measure?
 
  1. The ‘information burden’ on individuals and teams. “Certain people act as a lightning rod and everyone goes to them for all sorts of information. There’s potential for those people to get burnt out,” says O’Hanlon.
  2. Communication spread. “We can look at whether people work with a tight group or whether they spread out across the organization by spanning different functions. This could link to their role type and the communication style that’s suitable for that role.”
  3. Team alignment. “Typically, the culture of a team is made up of the way in which people interact. And the way in which people interact is shaped by the culture – it’s like the chicken and the egg. By looking at the actual communications that go on, it provides a quantifiable lens on culture. For example, are there individuals who are communicating in a totally different way to the rest of the people in their team?”
  4. Workplace structure. “Some organisations have workers who start and end their working day at the same time, all the time. There are peaks and troughs through the day that are very structured. Other offices are more chaotic. In addition, certain people have a work pattern that’s much less structured, so they work after hours or on weekends. They might have a lull through the work day itself. Being able to see that relative to others in the organisation can help to explain productivity levels and the way that people are working together.”
  5. Communication styles. “We’ve analysed millions of interactions and it’s clear that people have characteristic styles of communicating. It’s interesting is seeing shifts in that style. For example, someone has become less autonomous and is providing a lot of visibility to their manager in what they’re doing. That can be a sign that they are unsure of what they need to do; they might need some additional development.”

There are countless possibilities for how this information is utilised at individual, team and organisational level. For example, at the end of the probation period, managers can see how well integrated a new recruit is with the business and what their relationships look like. It can be used to support performance reviews. It can be used in retention plans, especially as communication patterns can potentially be predictive of attrition.
 
“We can see whether people are being proactive, how responsive they are, whether their communication style is very hierarchical or more egalitarian,” says O’Hanlon. “That in turn allows you to identify the future leaders, the people who are thinking of leaving and provide that kind of insight to better manage the workforce.”

#pb#
Diving deeper
McGrathNicol, a company that specialises in using and analysing large volumes of data to solve business problems, has recently broadened its focus from the fraud, risk and compliance-heavy financial services sector to the HR space. Kieran Earnshaw, a director at the company, says there is strong demand from HR professionals for more sophisticated use of data analytics.
 
“While the original demand came from the risk and compliance perspective, what we’re seeing is a demand from the HR community as well,” he says.
 
The McGrathNicol analytics tool – in contrast to the Revelian tool – does look at specific word usage. It applies text analytics in order to track:
 
  1. The ‘colour’ of an individual’s language and how it changes over time. This provides an indication of emotional feeling, state of stress and other indicators of emotional wellbeing.
  2. Interpersonal relations: By looking at some of the patterns between individuals, interpersonal issues can be identified.
  3. Signs of churn: Text analytics can look for indicators around possible employee churn where people are talking about interviewing and looking for other jobs.
 
“It breaks a stream of text, a chat or a message down into what’s called a bag-of-words, from which we can run sophisticated algorithms to look at the meaning of each of the words in that bag and then look at it within the context of the conversation,” says Earnshaw.
 
Clients receive a visual dashboard which sits across the apps and models: this allows users to visualise and explore the communication networks and patterns within their organisations – essentially who’s communicating with who – overlaid with the key characteristics of the communication. For example, positive or negative emotion, indicators of unhappiness, or indicators of potential churn. In this sense, it’s similar to measuring customer sentiment via social media channels.
 
“You can click on an individual and look at how their content and patterns have changed over time. Looking at changes longitudinally is interesting,” says Earnshaw. “An individual may have a particular style of communication that is slightly different to other people – that in itself is not all that interesting. However, it may be that their communication style changes and becomes, for example, more aggressive over time. You can start being proactive and pre-emptive in how you address that.”
 
In addition, interpersonal issues within a team can be identified, as well as employee/manager relationships which have become slightly toxic and worthy of intervention. However, McGrathNicol is not a full-service HR consulting firm; they do not make recommendations on action to be taken once data results are collated.
 
#pb#
Privacy
Some employers will approach such monitoring exercises with caution, wary of claims of being branded as Orwell’s Big Brother. Is that justified?
 
Many organisations are already scanning email and other forms of communication for inappropriate content and comment, and they have a clearly stated communications policy in place. Transparency is critical. A written policy needs to outline what’s appropriate and not appropriate in electronic communications. This policy needs to be disseminated to employees and the employer must be confident they have read and understood it.
 
Nonetheless, Revelian’s O’Hanlon says there are some misconceptions to clear up.
 
“Most people automatically assume that others are reading their emails. However, because we don’t look at content, the type of information we’re looking at is much more aggregate. Fears about looking into individual correspondence – while a company might have the right to do that – it’s not something our tools do. It’s really at that pattern level.”
 
It’s also important to note that consent must be obtained from employees. “From the privacy point of view, it’s consent-driven,” O’Hanlon adds. “Staff will know what’s going on. They will provide the consent for analysis.”
 
There are key pieces of legislation that employers must adhere to with any monitoring of employee communications: the Privacy Act; and state-specific workplace surveillance legislation.
 
Each state has workplace surveillance legislation. The rules are slightly different but some general principles are, in most states, there is legislation which deals with different forms of surveillance: traditional camera or CCTV, tracking surveillance, such as GPS for vehicles, and computer surveillance. NSW, for instance, has the Workplace Surveillance Act 2005. That legislation states that, in order to conduct computer surveillance, certain requirements must be met relating to notification to employees that this surveillance is occurring.
 
It’s important to note that employee records are exempt from the Privacy Act. But what is and isn’t an employee record is a grey area. Employers must be conscious of what they’re collecting and whether in fact the Privacy Act applies. They also need to aware that they may be capturing non-employee information – for instance, if an employee emails their wife. In those instances, the Privacy Act would likely apply.
 
McGrathNicol’s Earnshaw has also encountered privacy concerns from clients. He says two words are critical: intent and access.
 
Employers must be transparent about this monitoring; they must communicate to employees that there is reasonable intent behind gathering this information.
 
“It’s got to be for the good of the employees and the organisation. It’s not Big Brother spying on everybody,” he says.
 
As for access, this must be for a particular purpose only, and only by certain groups or individuals within the organisation – for example, the HR manager.
Regardless, Earnshaw is constantly surprised that, despite repeated reminders and warnings, employees still do the wrong thing.
 
“They’ve used a word being tracked by an internal filter and they’ve been prompted that they are being monitored and should be careful. And they carry on. It’s an interesting phenomenon that people are made aware, yet they continue.
 
“I’m not psychologist but I think people assume that the organisation doesn’t really have the capability or resources to truly monitor what’s happening. And in a way, historically people would be right in that regard. In extreme cases, some people actually want to be found out. I see that in the white collar fraud space – people become careless and a lot of times it’s intentional. They want to be caught.”
 
Only the start
HR is at the start of this journey, and most organisations are not yet leveraging this level of data analytics just yet. “In most organisations the first thing they’d know is an issue is raised with HR. Then retrospectively they might go back and look at retrieving specific emails that may be presented to them by an employee as evidence of an issue in the workplace, and potentially workplace bullying. This is looking into the future and getting more proactive,” says Earnshaw.
 
O’Hanlon agrees and suggests that one misconception remains: “This is much simpler to implement than employers might think,” he says. “There is an appetite for doing analytics in most companies but also a sense that big data equals big cost and big complexity. By taking on something like this, which is easy to use, you can get some of the benefits of world’s best big data analytics very simply.”
 
 
 

Recent articles & video

Australia's paid parental leave to reach 26 weeks by 2026

IT contractor gets 2.5-year jail time for swindling

Can 'provocation' be used as a defence in a workplace altercation?

Should an employer's religious views influence a dismissal decision?

Most Read Articles

Teacher sends 'Goodbye' message on WhatsApp group: Did she resign?

'Frustrated' worker blames 'understaffing' for aggressive behaviour

Co-managers challenge improper consultation process amid redundancy