This week’s podcast guest is Nigel Guenole, Director of Research and Ethics at the Institute of Management at Goldsmiths, University of London and a former colleague from our respective time at IBM.

Throughout the episode Nigel and I discuss:

• How the field of organisational behaviour has evolved over the last few years

• 
What the value of the field is to the enterprise and what some of the new opportunities are

• The capabilities that companies should build in organisational behaviour and how to do this

• Nigel's recent research, looking into emerging analytical methods, skills identification, and building a skills ontology

Support for this podcast comes from iPsychtec. You can learn more by visiting https://ipsychtec.com/

You can listen to this week’s episode below, or by using your podcast app of choice, just click the corresponding image to get access via the podcast website here.

Interview Transcript

David Green: Today, I am delighted to welcome Nigel Guenole, Director of Research and Ethics for the Institute of Management at Goldsmiths University of London, to The Digital HR Leaders Podcast. And a former colleague at IBM. Nigel, it is great to have you on the show.

Can you provide listeners with a brief introduction to you and your work? 


Nigel Guenole: Yeah, it is great to be here too, David, talking to you again.

So my background, I am an industrial organisational psychologist. So I work on measurement of individual differences, which is like the knowledge and the skills and the abilities that people need, to perform well in their jobs. I also do a lot of research into statistical methods and machine learning methods to make more accurate predictions, which is why I do a lot of work in workforce analytics. 


David Green: And of course as I briefly mentioned in the intro, we worked together at IBM for a few years. We were both part of the Smarter Workforce Institute and I had the pleasure of taking some of the research that you and Sheri did, out into the external market. During that time, with Sheri and Jonathan Ferrar, you co-authored, The Power of People. It has been five years since the book was published and given that a lot of the people that listen to the show are either working in, or interested in people analytics, what do you see as the main difference in people analytics now versus 2017, when the book was published? 


Nigel Guenole: Sure. So when I look at that book I think, what it did well was, it talked about more of the social side of people analytics. How do you set up the function? How do you build a team? How do you identify the skills? And I think that what we are seeing now is a much heavier focus on the analytics itself. So, how do you make sure you have the right data? How do you clean the right data? How do you try different machine learning and statistical methods? 
And there is a lot of talk in that book, when I look back on it, David, about storytelling and influencing. I think that caught on a little bit more than I would have liked it to have caught on. I think that when I have a look at a lot of the storytelling and influencing, that was probably already happening in HR before analytics, and I don't think you need analytics to do it. What I think you do need is much more rigorous analysis, so that you are a lot more confident in your conclusions and when the conclusions aren't rigorous, I think we need to temper the influencing and dial back the storytelling a little bit.

So I think I am starting to see a little bit of that kind of thinking catch on, in the bigger organisations at least, the banks, the technology companies, the pharma companies, the outfits that have those sorts of high-end technical capabilities. 


David Green: Certainly during the course of that five years, obviously we are both closely connected to the field, we have seen the explosion of growth in people analytics teams in organisations that had people analytics teams back in 2017, and obviously the emergence of people analytics teams in organisations that didn't have people analytics teams in 2017. So the field, as you said, is getting more sophisticated around some of the analytical methods they are looking at. You mentioned machine learning, which obviously is an area that you have done a lot of research, I think we will pull that out later in our conversation, when we talk around skills identification perhaps, and using skills as something to build a lot of talent management practices on, moving forward. 
But what I really am interested to talk to you about, to start off with Nigel, is around organisational behaviour and some of the work that you are doing. How has the field of organisational behaviour evolved over the last few years? And what are some of the key opportunities in this area? 


Nigel Guenole: So I think that with respect to analytics, how I am starting to see it evolve is, I think one of the big mistakes you can make when you are in HR analytics or you are a HR team in an organisation, is starting from scratch. Just building the models, trying to identify, what do I think the causes of engagement are? 
What do I think the causes of attrition might be? And then, you just start from scratch, you have a chat to people in your organisation, actually the literature is full of research on what those drivers are going to be, going back 40, 50 years. You are in a much stronger position if you start there, you start identifying the variables there and then you check whether that holds in your organisation. So I think that really needs to be the trend.

As it relates to organisational behaviour and HR analytics, the big developments that I am seeing are, a heavier focus on some of the psychological attributes that are required for performance. So what we used to see was that people would recruit against these knowledge, skills, abilities, and psychological attributes that are needed, and then they would forget about it and it would just in a folder somewhere and not integrated into how we might have development plans for those individuals, post hire. How we might move those people around the organisations to be better for them and better for the organisation when you are starting to see these self forming teams and agile working environments, some of this information is really important. 
And the next kind of change that I am starting to see is, rather than that just happening on the individual level. So David gets hired, Nigel gets hired, Nigel's performance, David's performance. We are seeing aggregates of those kind of capabilities. Team level personalities and team level capabilities and those linking to unit level outcomes like business unit outcomes, even up to organisational outcomes like stock market indices, accounting indices, and these sorts of things.

So it is a more holistic approach to the analytics, that is happening now. 


David Green: And I guess in terms of if you can understand drivers of team performance and you can help effect that, you are going to have a bigger impact than just looking at individuals. 


Nigel Guenole: Yes, that’s right. There are some jobs where the individual really matters, right? When you get up to those higher levels, the upper echelons of the organisations, but ultimately what we want to do is have strong firm performance. We want to see strong firm performance and employee wellbeing and satisfaction and career advancement.

But it is that kind of aggregated level that we are really driving towards with HR analytics. 


David Green: And obviously two of the big disrupters, technology clearly is one and increasingly has been one way before the pandemic, but obviously the pandemic has been a big disruptor for the last two years. How do you see that having an impact on organisational behaviour? 


Nigel Guenole: Yeah. There are some kind of challenges that organisational behaviour needs to start looking at, like a bigger focus on how we manage hybrid, how we might manage hybrid working environments, how we manage people when we are not co located. How do we get the benefits of being co located while still being safe. These are some of the things that we are starting to see in the pandemic. There are actually, a number of articles coming out in the technical and the academic literature. Actually, one thing that I would recommend that everybody do, all HR analytics directors, it is easy to do, just sign up to the table of contents of those top journals and then you just get fed into your inbox what the interesting studies are, what the interesting findings are, and those journals cost a fortune to actually subscribe to, but the table of contents are free. So I think there is a lot of useful information in there for HR analytics practitioners. 


David Green: And back to the initial point you were making around that, there is so much research that is being done on, as you said, things like causes of attrition, high or low engagement etc, there is so much to learn from what has already been researched over the last 40 or 50 years, as you said. 
So as a people analytics leader or people analytics practitioner, it is important to take that outside view in, before you just start, as you say, focusing on doing it within your own organisation. Yes of course you have got to check that what is being researched and what findings have been found, do apply to your organisation, but I am guessing if it has been done for 40 or 50 years, it is likely that it will do to a greater or lesser degree. 
It is a very key point, isn't it? 


Nigel Guenole: Yeah. It is a key point. And there are two schools of thought in the scientific literature, the organisational sciences. One is, that anytime you study the same thing, across different organisations and you get different results, for example you say the cause of attrition over here is salary, but the cause of attrition over here is job dissatisfaction. They would say, anytime you see differences across organisations, the reasons really are that there is something wrong with the analysis. It is low sample sizes, it is unreliable measures, or it is range restriction. And what they actually say is you should just trust the research.

So I am not quite there, I don't quite think that I would do that, but I certainly would start with, these appear to be the drivers of attrition across most organisations, that is our field of where we will start now. 
Let's explore how that applies in our own organisation. 


David Green: And of course, at the moment, we are seeing quite a lot of headlines around the great attrition or the great resignation, whatever we want to call it. I would love to hear your view, as an IO psychologist, that A] lots of these sample sizes seem to be quite small and B] they are focused on intent, rather than action. I would love to hear your comment around that and maybe for our listeners, explaining that difference between intent and action? 


Nigel Guenole: When we look at that intent versus outcome, we call it a turnover intentions. And when we talk about turnover intentions, there is a continuum of withdrawal from the organisation. So when things are starting to not be quite the way that we want them to be at work, we might just start turning up late, first of all, and then we might just start getting slow to respond to emails and not volunteering and helping out. And then we start taking sick leave and then the next thing, I am looking at jobs on my phone.

So that intention, I think it is something that we need to take seriously, but it is not quite the same thing as everybody actually having left their job roles.

Intention is often something we do try to focus on in IO psychology capacities, because that is something that we can influence, that intention, we can change that intention. But when the person is out the door, it is too late.

So sometimes focusing on that one step before, that attitude, that feeling that something is wrong, is not such a bad place to be. 


David Green: We are going to talk about some of your current research now. So I would love to hear about that. 
What are you currently working on? 


Nigel Guenole: Currently, I am working on a few different areas. The first one is with Podium, which is an assessment business and there I have been designing psychological assessments that predict work outcomes. And those work outcomes are what we call, task performance. So that is the quality of work that is done. The quantity of work that is done. Whether the work is done on time. Whether it meets the needs of your stakeholders.

Also what we call, citizenship behaviours. These citizenship behaviours as things like volunteering for extra work, when it is not formally in your job description, but if everybody did this around here it would be a good place to work. So we try to predict the people who are going to be like that and we also have assessments that will predict those withdrawal behaviours that we just talked about, which is the turnover intentions.

Doing a bit of work still like we were at IBM, in the artificial intelligence space, particularly around bias which is such a hot topic. With the bias, it is an interesting one. First of all, there is a lack of clarity around definitions of what people mean by bias. So when you just talk to the lay people, you talk to people in the media or you talk to popular accounts of bias, that just means whether there are differences in selection rates. So maybe women will get selected less than men, black people will get selected less than white people and that is considered bias. There are formal rules that say what is acceptable and what is not acceptable there. 
In the psychological literature, bias actually instead means, does this assessment or does this tool predict in exactly the same way for all groups? So you can actually have a tool that predicts in exactly the same way for all groups and we say, it is not biased, but you could still have that impact there. And there are adverse impacts against one of those groups. So there are ways that you can fix that, well there are ways that you can address that in your selection system and there is a computer science school of thought, and there is a psychology school of thought.

The psychology school of thought is, let’s measure something different. So maybe if we are seeing some differences on one kind of predictor, maybe that predictor is not one we will use, we will use something different kind of predictor.

But the computer scientists are much more adventurous with their methods and they will actually go in and they call it, data massaging, so they will massage the data just so that when they make the prediction it is fair. 
When I first saw some of those things, I just fell over backwards. Like I could not believe that this kind of thing was happening, but now I am warmed up to it a little bit more. But what I would say is, when you are working on bias, trying to reduce adverse impact, and if you have got computer scientists in there, put a psychologist in there or an HR professional, because the HR professionals are just aware of what you can and can't do. 
Sometimes I have seen the computer scientists try things and I am quite surprised, I know that the regular HR person, with a little bit of exposure, would never try it.

So I think you have to bring those two different schools together. 


David Green: That makes a lot of sense. And going back to, The Power of People, you talked about a people analytics team, one of the six areas or skills that you looked at was IO psychology and then another area was HR background. Wasn't it? 


Nigel Guenole: Yeah, you definitely need all of those skillsets in the team, or somebody that can cover a few of those. I don’t know what you are seeing as the size of organisational HR analytics teams now, if you take reporting to one side, because I think in the book, we were saying that maybe 6 to 10, back in 2017. 


David Green: Well we are definitely seeing that grow. In some of the companies as I am sure you know, particularly one that we may have worked in, you are looking at 70 plus in some companies. It is difficult to compare apples with apples because as you said, some include reporting some don’t as part of those teams. I know that reporting is getting more and more automated now, so you would potentially need less people to do that. But yeah, I know when we studied this in 2020, the average size of the 60 or so companies that we looked at, who are all 10,000 employees or more, was 14. 
These teams are starting to have more and more of a positive impact, both on workforce outcomes, but also business outcomes. And I think the pandemic has helped elevate these teams in many organisations as well because of the work they are doing, particularly when it is connected to employee listening and understanding how employees feel about suddenly being remote, or how managers are feeling about having to manage a team that they don't see every day. 
So there are lots of reasons for it, but I think it is mainly because it is having an impact. It is impacting on outcomes, which this is what, you, Jonathan, and Sheri, were writing about five years ago, about setting yourself up to do that.

One of the areas where we are seeing it growing, I know an area that you have worked on a lot and there is a lot of interest, is around skills identification and building a skills ontology across many organisations. What advice would you give to organisations looking to do this, based on your research? 


Nigel Guenole: The biggest lesson is similar to that one about, not starting from scratch. There are so many skills ontologies out there today. O*NET, is the big one that a lot of people will be familiar with, from the US department of labour. Maybe that is a little too US focused for some organisations, but there are also European skill taxonomies available from the European Union. So you just have to have a look at what kind of skills framework that you are going to use and then the big challenge is to embed it in all of the different HR processes.

Start from the attraction of the candidates. So you need to somehow communicate the knowledge and the skills that are going to be required in those job ads, but then you really want alignment across all of the other processes. So what you communicate at that point of the person coming in is, what you develop the person against, what you train the person against, reward the person against, performance manage the person against, and that is the kind of thing that, I think, you need to work very hard on to get that well embedded in the organisation. Particularly in large organisations and organisations where there has been a lot of M&A activity, you have got so many different competing frameworks and models, I think that the big challenge is standardisation and rationalisation.

One of the things that quite impressed me when I was at IBM with you, was their framework for changing organisational skills for strategic change. So what they managed to do was identify the skills that are required for where the business needs to be in 5 years, 10 years, and then they communicate that to every employee in the organisation. Then they offer all different forms of learning and learning management systems and they tag them with the skills, so that the employees know the skills that they have got to learn. And then there is a dashboard for the senior level HR execs and they can log in and they can watch the organisation transforming its skills as more and more people enrol. They can see the number of people that have enrolled in different courses, the number of completions, the number of dropouts. So that strategic use of skilling I think, is really powerful. 


David Green: And obviously IBM is one organisation, there are others that are using machine learning to really help industrialise or scale skills identification. Again, I am not sure if you have done a lot of research around that, but again, what would you advise companies around that? As you said, having that language across all the talent management practices is one, but any other advice? 


Nigel Guenole: Yeah. The way that a lot of those machine learning methods are doing it now, is they are inferring the skills from the digital exhaust, we call it, that you leave as you go about your work. So it may be in a public Slack channel, you have answered some questions that show that you have some skills in a particular database technology, or there is something on a corporate intranet, or maybe there is something on Github, or various different sources and these crawlers can gather that information, pass that information, score that information, and tag you as an employee inside the organisation as having those skills. 
I think that, what it is, is a decent starting point, but it is certainly not the end point. Because I think that there are people who have all sorts of skills but they are just not the people who are showy about it and putting it in all of those public places, whereas some people are really good at putting it in all of those public places. So what you really need to do is use that machine learning to start off to build your map of the people and the skills, but then let everybody update it. So then they can say, this is what we think we have discovered about capabilities just from the machine learning. Can you go in there and update and add any new data and add any new evidence in there? 
I think that is a more inclusive and likely higher quality result at the end of the day. 


David Green: And I guess that machine learning gives you that start. We have seen, when you just go out to all employees and ask them for their skills, you are basically doing it from a standing start, the machine learning gives you quite a big starting point. 


Nigel Guenole: Yes, it really can give you a good starting point. I guess the other big concerns, you and I wrote a paper about it, is the privacy side of things. Where if you are going out to external websites, you have to be careful that these are only professional websites, right? You don't want to be scraping anything social that is not work-related, or you don't have the consent to do. 


David Green: And I guess, again we talked about this in that paper and some of the other stuff that you were involved in at The Smarter Workforce Institute, having that clear value for the employee and ultimately understanding their skills and helping them with adjacent learning, or career opportunities within the organisation, or mobility. As long as you communicate that properly, you can see that there is a clear benefit, potentially, for employees, if you do that well and do it right. 


Nigel Guenole: That is right and I think that people are more willing to share information, if they believe that they are going to get something beneficial in return for it.

Transparency as well, this is another big thing. Tell everybody that you are going to be doing it and maybe give everybody the opportunity to update any information in those sorts of repositories beforehand, like LinkedIn, like Github, like Slack, any kind of corporate intranet and that sort of thing, before you do all of the data collection, that might be a way to go. 


David Green: Agreed. So next question, I am going to ask it on the three levels. What value does behavioural science bring 1] to people analytics, 2] to HR, and 3] to the organisation? 


Nigel Guenole: So for people analytics, it just helps you reach accurate conclusions and come to the best decision rather than erroneous conclusions and results that don't make sense. That is what it essentially does. And it can do that really well if you start from defining your research questions, do a little bit of a literature review. Google Scholar is something that people might not use, but Google Scholar is really good. A lot of that stuff is open access now, so it is really powerful and it can help you to frame the sort of analysis that you are going to do. Even some of the storytelling and influencing, the writers are doing that on their papers as well, because they have to convince the readers and the reviewers. So that is a really good resource to help you get to the right conclusions.

For HR analytics, I think that the analytics without the behaviour is not really HR analytics. I think that it is a pretty important component. For HR analytics, I think it can just help elevate the discipline. It can elevate the discipline if you get the right kind of messaging coming up. I don't know what you are seeing around where the director of analytics is reporting these days, I think at the time it was mostly into the CHRO but I think it can help them to report into the CEO more often.

So I think that it can elevate the discipline and just make it have more of a serious impact. I don't want to go this far, but look how seriously we took all of the scientists in COVID. I think that if we really major on the behavioural science and HR analytics, it will just elevate the messaging that we have there.

On an organisational level, it can certainly help drive organisational performance. There is no doubt that if you set that up right, you can show empirically, statistically, that these approaches, when you implement them well, lead to a greater organisational effectiveness.

And for that, David, I think that what you really want to be showing is causal inference. I have been talking to a lot of people about whether we can really make causal claims on the basis of our HR analytics analysis. And I think for a lot of the analysis we do, you can't, but we really need to change that. There is a massive demand for it, that is what the business needs, but I think that is a skill shortage area as well this causal inference area.

So that is what I would be hoping to see, in the coming kind of three to five years from HR analytics. 


David Green: Okay, quick follow on one around reporting lines. We are seeing an increase in heads of people analytics reporting directly to the CHRO. In the research we did last year, this is for the listeners as well, 22% report directly to the CHRO and then just over 75% either report direct to the CHRO or to someone on the HR leadership team and that is trending upwards. So I think really important, if I was CHRO, I would want my head of people analytics reporting into me for the reasons that you gave. 
And then the second thing around trying to show causal inference, what are some of the tips that you would give to heads of people analytics and other senior HR leaders around how to measure that? 


Nigel Guenole: The big thing is it is research design. So it is not any kind of statistical analysis, it is not any kind of machine learning analysis, but it is research design. And research design for me, is questions like where am I going to collect the data? How am I going to collect the data? When am I going to collect the data? Not what analysis am I going to do on the data? And if you can answer those questions well, about what data will I collect? 
And where will I make an intervention? And you get the timing right, you can start to make those causal claims.

The strongest approach that you can do is these controlled trials, like they are doing medicine, but we are not in medicine and we can rarely just randomise people to different conditions. But there are techniques that the econametricians use and the epidemiologists use, and we can start to borrow those and bring those into HR analytics.

I have seen a couple of papers in review processes, starting to talk about those topics so I think it is going to emerge as a hot one. 


David Green: So it could be one for us to talk about on the second time we do a podcast, yeah?

So then obviously the way you get to this is through having this behavioural science capability within your organisation. What is your advice for companies that are looking to build this capability in behavioural science or organisational psychology, how can they do this? 


Nigel Guenole: One way is to hire an organisational psychologist. To have that as one of the roles in the team. If you have got a small team, maybe that organisational psychologists can double as the analytics person. I think that the modern IO psychologist is not just good on the theory, but they are also quite strong on the analytics, they will hold their own in those kinds of conversations, with pretty much anybody. 
So I would hire that. And when you are hiring, just look for somebody who is hungry for it, that has the passion for it. I think you can spot it, as the people who are always thinking about that topic, familiar with the latest research topics.

So that is what I would do, hire one in. 


David Green: And actually, even if we go back to, The Power of People, two of the people you interviewed who were people analytics leaders at the time, got bigger roles now, Alexis Fink and Thomas Rasmussen, both IO psychologist, both really are seen as two of the leading lights in our field. 
And then I guess another way you can do it, if you are a smaller organisation is maybe buy some of that expertise in as and when you need it. 


Nigel Guenole: Absolutely. So there are many smaller consultancies doing that kind of work and there are many independent IO psychologists. So you can bring them in on shorter term contracts to do that sort of work. And then the thing that Insight222 is doing, all of that information sharing and all of that networking. So identify any IO psych people in that network and befriend them and find out what they are doing in their organisation. 


David Green: Can we break that IO psychology skill-set down a little bit further? What skills can people develop without professional re-training? And, when does becoming an accredited psychologist really become necessary? 


Nigel Guenole: So I would say that the analytical side, that a lot of industrial organisational psychologists do, you can learn that. There is incredible learning available on Coursera, edX, Khan Academy, these different MOOCs courses, the massively, what is it online? I don't even know the acronym, David.

David Green: Massive open online courses, is that it?


Nigel Guenole: There is incredible courses on Python and this kind of thing, so on the analytics front, I would say that you can do all of that. There is a lot of knowledge on the more psychological side, that might be harder to acquire, but you still can acquire it. I would say that the component that is really important, to almost learn more formally, as the ethical responsibility of what it means to be a psychologist and what it means to make a decision about a person and things like the information you should be sharing about your decision that is procedural, fairness, and justice, and all of that kind of approach, there is a qualification that you need to have in the UK at least, from the health and care professions society. So you can't just call yourself an occupational psychologist here in the UK, it is a protected title.

But you can learn a lot of that kind of stuff, but you have just got to learn where your professional boundaries are. What do I know? What don't I know?

So my professional boundary for instance is, I do a lot of work and measurement of maladaptive personality, but I would never touch clinical psychology because this is not a specialism for me. 


David Green: Yeah, that makes sense. And as you said, there is increasingly courses available online and some of them are free, some of them are not very expensive. So you can certainly pick up some of those analytical skills that a psychologist usually possesses, or should possess. So yeah, very interesting. 
Nigel unfortunately we have got to the last question now. This is the question we are asking everyone on this series, how does behavioural science help to improve the workplace? 


Nigel Guenole: I would say we have talked about how it improves firm performance and I think that is pretty clear.

One of the big contributions that it can make to improving the workplace, if I am to interpret that a little more broadly, it is just by helping make work more meaningful. Helping the work become more meaningful for employees, by ensuring everybody is healthy and so far as possible, enjoying work. And we know a lot about what those things are from different models, like different stress models, what are the causes of stress and how can we ameliorate or mitigate those causes of stress. We know a lot about how to design jobs so that people find the jobs enjoyable or manageable.

We seem to be in a bit of a sweet spot, David, if I can call it that. We had what we called taylorism, back in the early 1900’s, where they used a lot of new technologies. So instead of us going to work, the work came to people and they just made us work as hard and as fast as we could and it was not a lot of fun. Then people found out that people got sick because of that and they weren't as productive as they might be, if we designed jobs more effectively.

So now, we are in a place where, some jobs are dangerous jobs and not fun jobs, but for a lot of the white collar jobs they are not too bad compared to where they are historically. But the risk is with machine learning, where we are just chopping up the jobs into these little pieces, taking them away from people, and the humans doing what is left, is we are at the risk of the meaning disappearing from work.

I think that some of the organisational behaviour theories, can help us design the jobs that include artificial intelligence and machine learning, so that they are still rewarding because we spend a lot of time doing this, don't we? 


David Green: We do. We do spend a long time, most of our time in fact, apart from probably being asleep and some of us spend more time working than we do being asleep.

There was one question that I honestly missed and I know it is one that I think our listeners would love to hear your thoughts on. Do you have any examples of companies who are building a capability in behavioural science particularly well, that you are able to share? 


Nigel Guenole: Let me have a think. I really only know the big ones. There is IBM and these kinds of companies, but one is Ezra. Ezra is a coaching company that is doing some pretty remarkable things. I am doing some work on those guys and they have some great analytical capability internally, where they are looking at the impact of coaching on individual performance, but also organisational performance. They have got some great methods for studying change in there. 


David Green: And as you said, linking that back to creating meaningful work. There are a lot of studies out there showing how that has such an impact on people's performance, their health, their engagement, their productivity. So it is actually a really important thing, this whole concept of meaningful work. 


Nigel Guenole: Absolutely. Yeah. It is the big topic of the coming years. How do we retain that meaning in work, while integrating technology. What is that land going to be, of the technology and the human. 


David Green: That is a great place to end our conversation, Nigel. As ever, always enjoy talking to you, thanks for being a guest The Digital HR Leaders Podcast.