HR analytics Interview: Patrick Coolen
On September 14th, I interviewed Patrick Coolen. Patrick is a recognizable face in the HR analytics field and I was very curious to learn about his experiences in applying people analytics at ABN AMRO. We talked among other things about better decision making, business cases, most important lessons learned and his vision on the future of people analytics.
Hi Patrick, could you please introduce yourself?
My name is Patrick Coolen. I am the head of HR metrics and analytics within ABN AMRO. I have been doing this now for three years.
I studied management of information at the Tilburg University and started my career at ABN AMRO within IT. After three years working in IT I moved via recruitment into HR and never left. Since then I worked on several strategic projects and was part of the MT HR before I started my current department.
I was always passionate about HR activities like performance management, succession management, talent management, et cetera. But the longer I was in HR, the more curious I became on investigating the actual impact, or return on investment if you will, of HR on our business.
So what really happens when engagement increases to products sold or what is the actual effect of developing a specific leadership style on client satisfaction or net growth? Answering these types of questions demands a more fact-based or scientific approach. That’s why I started at the department HR metrics and analytics.
Related (free) resource ahead! Continue reading below ↓
People Analytics Resource Library
Download our list of key HR Analytics resources (90+) that will help you improve your expertise and initiatives. Your one-stop-shop for People Analytics!
How did you get into people analytics?
Patrick: One reason is a little bit of what I just said about curiosity and wanting to know how the impact of HR actually works. The main reason for me is that our business is simply asking different questions than a few years back. It is not enough anymore to tell our business that engagement moved up with 2 points or that absenteeism went down with 1%. Do not get me wrong, this is important, but knowing the effect of these changes on engagement, absenteeism, sales, client satisfaction or other relevant business goals really gets the business interested.
Another important reason is that it can be risky to make decisions without using advanced analytics. Using statistics or machine learning methods simply decreases uncertainty in predictions compared to only looking at for instance a pivot table or graph.
The fact that two numbers are unequal does not mean they are really different and it certainly provides no information on why the possible difference is there. For instance, when we see that the average pay for men and women in a specific role differs, our first reflex is to have a discussion on gender equality. Advanced analytics, that normally is able to take more people factors into account, may reveal that gender is not the issue but for instance education or having a specific competency. So advance analytics helps to find the actual root cause or drivers if you will.
Erik: I came across something very similar when researching employee innovativeness in a professional service firm. The analysis showed that men were more innovative compared to women. However, further analysis showed that most of the partners within the firm were men. They had the authority to come up with innovative ideas, promote them and implement them in the organization. The women were overrepresented in the non-partner ranks. They had less autonomy to display those behaviors. So you see it happen for performance, but also for innovation.
Patrick: Exactly, that is a similar problem. I am passionate about having a smarter look at data (in other words: using advanced analytics) in order to decrease the risk of making wrong decisions. The best application of advanced analytics is for us typically in the business itself. Business goals like client satisfaction, sales, growth, market share, costs or quality of work are good examples for HR analytics.
A lot of people are talking about HR analytics and a lot of people are talking about people analytics. Which one do you like best?
Patrick: That is a good question (laughs). I am using people analytics more and more. HR analytics has the connotation that it is about the HR department. This is partly true, but in many cases you are looking at people characteristics that are not managed by HR but by the business.
People analytics reflects more the research we do when we look at engagement characteristics, or capabilities characters, or team characteristics, or leadership characteristics – and how they influence the business, like products sold, net promotor score, client satisfaction and other outcomes. These are all people characteristics. That’s why people analytics is a good word for it. But I use them both.
I do get the question more and more. There is also the term advanced analytics! Just make sure that you explain what you mean whenever you use one of these terms. I always say something like: “People analytics is using statistics or some data mining technique on combined data sets of HR and the business to find relationships in the data that improve decision making”
ABN AMRO has been one of the leaders in fact-based HR. You once said in an interview 3 years ago: “It is our mission to become an HR organization in which we want to support decision making in a way that’s fact-based”. Are you there yet and what obstacles do you still need to overcome in order to get there?
Patrick: We have come a long way – but we still have ambitions. We are very satisfied with how things are going. That means that we are doing more research on more data sources on multiple business targets and service more business units than three years ago. That is the progress we’ve made.
One ambition we have is to improve on action-ability, and supporting the use of insights. Our business is committed to take over our research outcomes but it takes focus and discipline to periodically evaluate progress over time. We are a relatively small department and do not have the resources to support business on this topic as much as I would like.
Another thing to think about is how we are going to move from HR/ People analytics to enterprise analytics. Is an enterprise view on analytics beneficial in terms of economies of scale, sharing data et cetera? And of course, always on my list is to continuously improve on the use of different analytical techniques and analytical software.
Josh Bersin once said in an interview with Luk Smeyers: “It takes 5 years to get to the highest [HR analytics] maturity level”. Do you see that in ABN AMRO as well?
Patrick: I think that’s a good estimate. We are growing every year.
Like I said, we are using more data sources and more methods of data mining and machine learning and using different tools and keep up with the different tooling – which is almost a day job. So we are getting more mature year by year. But for me the highest level of maturity is moving to enterprise analytics like discussed in the previous question. Transforming all the pockets of analytics in our organization towards an enterprise vision on analytics will take about two more years in our organization.
So what you’re saying is that HR analytics is its own undoing. If it becomes successful, it will become enterprise analytics?
Patrick: That can be a scenario. But in that scenario people analytics will not disappear but it will become an integral part of enterprise analytics. It is more about how you organize data, what IT-platform do you use, do you use cloud-based, on premise, centralized or decentralized IT platforms, is it making all data accessible from one point or multiple points, or can we keep data separate? Where do your data scientists go? Is that an external or central team, or can it also be decentralized? That’s all organizational and IT design. The best solution depends on where you are as a company.
Can you give an example of how analytics lead to better decision making at ABN AMRO?
Patrick: Well, in all our research we find relationships that for 70 to 80% confirmed the strategy we are following. That is true for our business as well as for our engagement strategy, our recruitment strategy or for instance our diversity strategy. Although these relationships are intuitive they are already helping to focus on what ‘buttons’ to push. This because intuitive relationships can be ranked by impact strength. The other 20 to 30% deals with more counter-intuitive relationships or insights that help you to change or enhance your strategy or intervention.
Would you say that analytics lead to HR efficiency?
Patrick: Analytics improves business efficiency and it improves the effectiveness of HR activities. These relate to each other. More effective HR activities should lead to better business results.
I am asking because some analytics enthusiasts I talked to, speak about some of the more traditional HR activities become obsolete when an organization starts to make fact-based decisions. Do you recognize this?
Patrick: Which ones? Because of analytics leadership is not disappearing. Because of analytics learning and recruitment are not disappearing. I mean we are making these areas more mature and effective. Analytics can support the recruitment process and reduce a long list of candidates to a shorter list in a more effective way, and it can help us sharpen the profile we are recruiting on – but in the end you need a recruiter to bring the data together and make that decision.
I don’t know if there will be an algorithm where information will be put in and then the decision is made whether a candidate is hired or not. At least we are not there yet. But it will affect the work in those areas. It will not make the recruitment function itself obsolete.
It will make the job of the recruiter easier because he doesn’t have to scan through all the CVs. But doesn’t that also mean you need less recruiters. Do you need less people in order to do the same work more efficiently?
Patrick: I expect that analytics will change how you do recruitment in your example and consequently it demands different skills from a recruiter. He or she should at least be more data savvy and curious to learn what all the recruitment data tells us.
What are the most important lessons learned for the past few years at ABN AMRO?
Patrick: I think one of them is: stay close to your business problems. That is the strategy we use since the beginning. We wanted to show the additional value of HR analytics as soon as possible. For us that meant staying close to a real business problem. Don’t start investigating attrition if it is not a real problem for your business. Don’t start diving into engagement if it is not a big issue for your business but start talking about client satisfaction, quality of work, or commercial KPIs like sales or net growth.
Make sure to have a discussion with your business to find concrete opportunities or problems – and start determining how people variables might impact those problems. Then you’ll have the attention of your business.
Also try to focus on action ability. You have to make sure that your business can take action on the outcomes of your research. For example, if you find a relationship between satisfaction on leadership and sales you really want to know what is driving leadership ‘satisfaction’. Only than insights get actionable.
Privacy and legal is a third important factor. You are dealing with employee data, so make sure you have a close line with the legal department.
Another advice I want to give is: do it! Maybe it looks difficult when you are not a statistician, or a data miner, but you can get help with that. Start experimenting a little bit, start small and start tomorrow.
I am a little bit puzzled why most organizations are still not applying analytics on people data. I also mention this in the introduction of my latest LinkedIn post on the “10 golden rules of HR analytics”. When you look at the Bersin report you see that the number of organizations who think they are fully capable of doing HR analytics has increased from 4 to 8%. You can look at that and think: “hey, it doubled!” I am thinking: “what is holding back the other 92%?”
By far the largest part of all organizations think they are not fully capable. 70% is not even feeling somewhat capable of doing HR analytics.
I think you should stop thinking about it as a very complex thing. Yes, you need specific skills including statistics, data mining and domain specific skills. You also need some business knowledge. This is all knowledge you can organize together in a team or per project. Don’t be afraid to play with data and experiment a little bit. Don’t look at it (only) from a risk perspective, but see it as an opportunity
Isn’t the HR community also kind of guilty of portraying analytics as something really sophisticated, involving data mining and all kinds of difficult algorithms that nobody in HR has ever heard of, while you can start really simple by combining two separate data sets and do a simple correlation analysis just in Excel?
Patrick: I cannot answer that for the whole HR community. Maybe you are right. I always say that you can start tomorrow. I try to make it less complex.
“HR is in nature more relationship, change and development driven than data driven. I truly believe they go together.”
The point is that up front you can look at all the obstacles like legal, data, algorithms – or you can start small and have more of a ‘just do it’ attitude. I think in general HR is a bit afraid of number-crunching, instead of embracing it.
What is the biggest change you have seen happen in the last two years in the field of HR analytics?
Patrick: In general, like I just said, I am still surprised that there is 92% of organization still not capable of doing HR analytics. So I don’t see enough change. I am actually interested in why organizations are not picking it up?
When we started we were ‘selling’ HR analytics directly to the business. Now the demand is coming more and more through our HR organization. Where three years ago maybe 5% of our research portfolio was initiated via our HR department, it is pretty much 50/50 now. That is a good thing. It means that our HR department is recognizing the opportunities That is a big change because fact-based HR is not only about my department. It is about the entire HR organization.
What are you most excited about in the next few years?
Patrick: Two things. One is what we already talked about: how can we benefit from a more enterprise vision on analytics: how can we bring that to a next level within our organizations. This is a capability discussion within HR on the one hand and an organizational design discussion within the enterprise on the other hand.
I am also excited about stepping up with regards to methodologies, tools, data visualization. I have a strong appetite to keep on learning and improving our ‘game’. The urge to grow and learn has not stopped and as long as that is the case I am extremely happy to work in HR analytics!