The Future of People Analytics

I make no secret of my interest in predictive behavioural analytics - understanding why people do what they do, and whilst I often talk about Customer Experience (CX), I am just as interested in Employee Experience (EX).

Many experts in the field of Customer Experience also stress the connection between Employee Experience and Customer Experience, and there is plenty of reliable research to prove the business value of that relationship. When you understand 'why', you can influence 'what'.

However, I have noted that compared to other disciplines within business, People Analytics (aka HR analytics) still has plenty of room for improvement - both in capability (see my blog about analytics evolution) and in deployment scope. A data platform for people analytics tends to be insular - focused on data about employees, their performance or the progress of specific process (see my comment about ATS below). Most of the available packages seem to rely on operational reporting, with some basic diagnostic capabilities and a few use this to support workforce planning and resourcing. Few seem to offer any significant capabilities to understand what is driving those measures and almost none offer the ability to optimise labour acquisition and deployment or support managers in making decisions about their teams.

In this article I will discuss 5 key employment stages where innovative solutions for people analytics can play a critical role, as well as on a particular aspect of employee experience - human behaviour.

Introduction

In a blog I wrote some time ago, I discussed the importance of understanding how emotions contribute to human decision making and the resultant observable behaviour. Here is what Herb Kelleher says about hiring and attitude (one manifestation of emotions):

"You don't hire for skills, you hire for attitude. You can always teach skills."

So popular is this leadership philosophy that there is a bestselling book on the subject - Hiring for Attitude by Mark Murphy Attitude might be considered as a stable set of emotional preconceptions that influences the mercurial world of emotions (as well as being subtly influenced by them). Attitudes can predispose us to a particular set of behaviours or determine the decisions we make.

Herb's point was that without the right attitude, employees can not or will not deliver their full potential. I would argue that this goes beyond an employee's ability to learn or to perform well, and touches every aspect of their working life - from before they join an organisation to well after they leave.

But what is 'attitude' and how do you measure it? Furthermore, can you keep doing it through the lifecycle of an employee, and is it worthwhile to even try? I am convinced that it is worthwhile, even vital, that we use new analytical techniques to understand what makes our employees happier, more confident, productive and loyal - even more so in the current uncertain climate of coronavirus. Note that I said 'new analytical techniques', I do not mean the plethora of basic reporting tools that seem to make up the bulk of current people analytics solutions today.

If you want to understand why employees do what they do, you need to listen to the 'Voice of the Employee' - their authentic voice, through which attitudes, beliefs and emotions are expressed - because these drive behaviours and outcomes. Typically, this does not come through surveys, quarterly manager reviews or formal channels, but through the informal narratives that make up their day-to-day conversations with co-workers. How you gather and analyse that data will vary from organisation to organisation, but it can be done.

But before I go into the 5 stages, I have some words of caution:

• To get the best out of analytics, especially advanced analytics, you do need to have a reasonable understanding of data and data analysis. Experience shows that deploying analytics and automating parts of a decision-making process without this experience can lead to significant problems downstream. As a colleague once told me 'a fool with a tool is still a fool'.
• Avoid fixating on a small number of data points and KPIs - as with customer experience, it is as important to understand what is influencing the result as the result itself. Despite often being described as 'our most valuable asset', humans are complex, messy and unpredictable - part of our role as leaders is to accommodate that and build on it. If we treat employees like 'meat machines' we can hardly expect them to 'go the extra mile', especially when the going gets tough - we are all so much more than cells in a spreadsheet or a line on a balance sheet.

Talent Acquisition

In a recent conference, we were discussing success factors for digital transformation; a common thread that kept coming up was the scarcity of talented employees. Other pundits talk about 'the war for talent' and 'the shortage of labour' and about the changing expectations of employees.

In addition, organisations (and would-be employees) are also having to cope with the changing face of work - with more emphasis on soft skills, as well as creativity, problem-solving and adaptability.

Whilst experience and qualifications are obviously still important, these are beginning to contribute less to an employee's long term contribution than these 'new' skills. Job descriptions (and resumes) need to change to reflect this, which makes the recruiters job harder; how do you specify, look for, and test for attitude without an overly intrusive (and often unreliable) barrage of tests? We've gotten so used to thinking about the quantifiable attributes of a candidate we've overlooked their behavioural potential. If we keep recruiting in the same old ways (albeit with some automation) we are in danger of building an organisation fit for the last century.

Analytics can help by giving insights into what makes an 'ideal' employee in a complex environment, or better yet, an ideal team. Once we know what these characteristics are, how they are exhibited, and where they are likely to be found, we can start to better target recruitment campaigns. The goal is to improve the quality, not the number of applications, or necessarily to just reduce the 'cost per application / hire'.

Another cautionary word; AI-based Applicant Tracking Software (ATS) and talent screening tools, whilst helpful, are not as smart as we would like to think; they work best for a well-defined, limited scope, but tend to result in a very homogeneous set of candidates; great if you want all your employees to come from a single cookie-cutter, but not so great if you want to promote diversity of talent and difference in points of view; creative leaps rarely come from a culture of 'group think' based on a single background or set of values.

Onboarding

Many leading organisations start the onboarding process well before the candidate walks through the door on their first day. Automation can play a valuable role in orchestrating the lead up to joining; scheduling content, interactions with managers and their new team, answering questions and reducing the workload. But the automation needs to be smart, responsive and able to tailor the joining experience to the needs of each new employee.

On joining, it can accelerate the journey to competence and productivity by enhancing initial training, providing opportunities to rehearse the employee's duties with smart, interactive training and role-playing that is tailored to the individual's strengths and needs. It can also anticipate their requirements, get them to the resources and support they need more quickly.

Similarly, analytics can identify where the onboarding process may need attention - addressing unmet needs or challenges in content and method by evaluating informal feedback from new joiners and their actual performance during their early deployment.

Our objective should be to offer a tailored onboarding process that can offer just what each new hire needs, bypassing what isn't and offering more support where appropriate. For some, that will mean an accelerated process, for others a slightly more intensive program.

Training

For both initial and ongoing training, analytics can support meeting the capability requirements of the organisation, as well as the learning needs of individual employees. Used in combination with a Learning Management System (LMS), it can also identify when a section of training may need some attention - whether it be the learner, the delivery method or the content itself. In addition, it can calculate the ideal timing, duration or intensity of training required (there is evidence that both too little or too much training can negatively impact learner motivation), as well as quantify the 'training decay' caused by any gap between learning and doing.

In the new 'Industry 4.0' era, with large-scale and pervasive automation, greater emphasis is now placed on humans engaging in creative knowledge-working with 'micro-learning' or Knowledge Management Systems (KMS) supporting on-the-job performance. With new tools for predictive behavioural analysis, soft skills, emotional intelligence (EQ) and underlying attitudes can also be evaluated and addressed, as well as knowledge and skills. Furthermore, Training is both affected by, and influences employee mindset and corporate culture.

Perhaps more importantly for those who commission training, is the ability to link training effectiveness to business outcomes; taking into account culture and employee mindset. No longer do we need to reply on 'happy sheets' (course evaluation forms) or potentially biased anecdotal reports to understand whether the learning opportunities on offer are meeting the needs of the organisation.

Engaging / Motivating

In my opinion, much of what I have written so far is widely understood (if not practiced). Some of the tools for talent acquisition and training do embed analytics (including Artificial Intelligence) but are primarily focused on automating a process; whether it be recruitment or training or scheduling quarterly reviews. In these areas, there is significant scope for standardised, modular steps - where analytics can provide the insights supporting orchestration. However, for many organisations there remains a significant challenge, especially in tumultuous times - maintaining employee motivation and engagement, which contributes to productivity and outcomes. The challenge is that there is no such thing as a 'standard employee' - we are all different, with different experiences, hopes and concerns.

Late in the 1950's Frederick Herzberg published his two-factor theory of job satisfaction (aka the 'motivation-hygiene theory'). In his theory, Herzberg states that there are independent causes of satisfaction and dissatisfaction, and that one is not the opposite of the other. Many of the factors that contribute to job satisfaction are emotional; being recognised for our contribution, personal growth, working relationships and so on.

Up until now, we've tried to measure these using formal research techniques (e.g. employee engagement surveys), line management and peer reviews, etc. Unfortunately, we know that these provide limited insights into what employees really think and feel - partly because of suspicion about how the data is collected and used (despite declarations of anonymity) and therefore a reticence to be candid, and partly because it is very hard to access our own subconscious. As I am fond of saying 'if you want to know how somebody is feeling, don't ask them how they are feeling'.

However, tools do exist to gather informal narratives (those 'watercooler moments') and surface both what employees truly value and why. Add that to operational data available elsewhere in the business and data about the employees themselves, and we can develop real insights to into what is driving employee engagement. In my experience (and this works with customers too), it's usually a combination of rational factors (pay, hours, family commitments, etc.), emotional ones (stress, lack of advancement, cultural dissonance and so on), internal biases and external influences, as well as prior experience and what's going on right now.

Convert these individual experiences into an ongoing employee journey and you have not only a good predictor of engagement and productivity but also of when an employee might be entering the 'danger zone' and considering leaving.

Retaining

In the past, I have been asked by organisations 'why do we lose some of our most talented and valuable people, especially as we offer industry-leading pay and benefits?', and there's an oft-quoted phrase "people don't leave bad companies, they leave bad managers!" I think that is a simplistic point of view. Yes, a bad manager can definitely contribute to an employee's decision to leave, but it is neither fair or accurate to blame managers for staff attrition.

As discussed in the previous section, there may be several causes of job (dis-)satisfaction, as well as purely rational needs; satisfaction is one of several factors that might prompt an employee to leave. For example, in one organisation I have worked with, the birth of a second child is known to correlate to an uptick in resignations - but why? is it the number of children? their relative ages (e.g. is the first child approaching school age)? If you know the why early enough, you may be able to take a proactive step to pre-empt and avoid that unplanned departure. What's more, you can assess whether it is better for the business to let them go, or to further invest in them - before the decision is taken out of your hands.

In addition, a more holistic view of business needs, team and individual capabilities can also give you insights into where you are strong and where further investment in people is needed - giving you the flexibility to deal with the loss of a key employee or changing circumstances. Succession planning is not just about having people with the right knowledge in reserve; besides you can't plan for every eventuality. Right now, around the world, organisations are struggling to cope with the impact of coronavirus and the need for isolation or social-distancing. Many of these organisations have disaster recovery plans in place to deal with unavailability of physical infrastructure, but fewer had anticipated the impact of not being able to co-locate their teams.

Conclusion

More organisations are adopting People Analytics, but the analytical tools they are using need to offer more than reporting of metrics or basic diagnostics. Even the adoption of AI in point solutions like ATS needs careful consideration and ongoing support and tuning.

Done well, advanced People Analytics can enhance recruitment, streamline the onboarding experience, optimise training & development, enhance employee engagement & motivation, and reduce staff attrition. Done badly, it can hamper diversity, focus on measures not outcomes, erode trust, and fuel a disfunctional culture or poorly performing teams.

Postscript - Offboarding

Finally, some organisations also care about what happens to the employees they do lose - whether it be about the 'Glassdoor' effect that disgruntled ex-colleagues can have on their reputation, or in leaving the door open for a re-hire further down the road.

If you are going to do exit interviews, make sure that you use the data to gain insights into why people leave, what might have been done to prevent it (if warranted) and what it might take to 'win back' key people. The answers might surprise you - more often than one might think, the answer is not 'more money', but something relatively small that was simply overlooked or undervalued. What really matters to people is not always what they say matters or what we think matters, and is often well within an organisation's ability to deliver.

If you would like to share your thoughts on this subject, please leave a comment and if you would like to contact me to discuss any part of this article - you can do so here.