Mike West is a People Analytics professional with his own company, PeopleAnalyst. He is based in Austin, Texas and has more than 10 years of experience in the field. Mike is currently working with clients such as Jawbone (wearable technology) and Pandora (online music). He was one of the first employees in Google’s People Analytics team and has made major contributions to People Analytics efforts at Merck, PetSmart, Otsuka and Children’s Medical. One could say that Mike is a geek who loves both business and behavioural science, he professes: “People Analytics is my reason for being, my life purpose – there is no second option for me.” Mike blogs about People Analytics on Tumblr and LinkedIn. He tweets (@mikecwest) and maintains a series of various resources that can be found here.
Johann: Mike, first things first. Define people analytics?
Mike: You can picture people analytics as the intersection of human resource management, behavioural science, mathematics and technology. If you don’t get Venn diagrams, that is okay too.
To summarise, “people analytics is the application of maths and science to human resource management”. This is easy to say and remember, but it doesn’t capture the complete essence of it. A better definition is “people analytics is the systematic application of statistics and behavioural science to human resources to achieve probability derived business advantages”.
The purpose of people analytics is to identify, produce and defend differences in people that matter to a business. People analytics are designed to solve people-related problems, make better people-related decisions, make better decisions regarding policies, practices, processes and systems, and optimise chosen outcomes within a business context. There are varying ways this purpose can be accomplished, but how people analytics is distinct from its predecessors is that it is systematic and it is oriented in a probabilistic view of the world, as opposed to a deterministic. view of the world
“On any given day we may still make mistakes, but as a whole we make better decisions and this adds up to real advantages…”
People analytics feedback loops produce new learning that leads to better investments in time and resources i.e. we reduce waste. On any given day, we may still make mistakes, but as a whole we make better decisions and this adds up to real advantages when distributed over large employee populations, over long periods of time.
Johann: Some would call you a data geek with a liberal arts degree. What brought you into people analytics?
Mike: As an undergraduate in college I was fascinated by the application of science to people. I was absorbed first by sociology – Emile Durkheim and peers – and I’m taken in by the idea that sometimes you can explain individual human behaviour better with group or situational variables than you can with individual variables. In exploring this, I fell upon social psychology, which you might think of as the operating system of the human brain – the connection of the autonomous individual to the whole. I took every class from the perspective of both Sociology and Psychology, ultimately obtaining a dual degree.
I was raised by a small businessman, in a broken family. I had two options. Go to graduate school or work in a dead-end job for the rest of my life for less than the cost of living, with no health insurance. On the other hand, nobody aspires to grow up and go into Human Resources. I went on to get a Master’s in Human Resources as a practical way to find meaningful employment to pay the bills and the student loans following a behavioural science undergraduate degree. I figured, in the end, business plays a very fundamental role in society and in human lives. It supplies the money, and it encompasses much of the human problem, and it also has the data I want! No matter how difficult it has been to do what I want to do in business, it is an unfathomable joy to me to get paid to apply my passion at work every day. I think everyone should.
I was lucky. My first job was at Merck, the pharmaceutical company. They were big, they had a lot of people, they were a very sophisticated company, and they practiced rotational leadership. A marketing executive rotated in as leader of HR and was mortified with how decisions in HR were made without the backing of data and defensible science, and that there really was no methodology to underpin decisions. This didn’t sit well with her, so she made changes. This was a major culture shift and it was hard going, but over time this made a huge difference at Merck. I too have rotated to different areas of HR, but the utilisation of data in HR has been my focus and the central unifying theme of my career since Merck. Every new role I have taken I have asked myself the question “what did we learn and how can we do this better here, knowing what I know now?”
Johann: You were one of the few people involved early on with People Analytics at Google. Can you describe what it was like to start in this field and what has changed since then?
Mike: I joined Google about a year before their official People Analytics team was formed. At the time, there were 3 or 4 of us working on this out in sub-function HR silos, sprinkled across various parts of people operations – what Google calls Human Resources. At the time, Todd Carlisle was doing fantastic work in Staffing Analytics and Sue Wuthrich, Head of Benefits, wanted a “Todd of her own”. I didn’t have a PhD like Todd, nor did I have the traditional Stanford or Ivy League education of most Googlers hired at the time, however, the conversation with Sue was very easy given my prior work in the space and it wasn’t long before I received an offer.
Within a year and a half of working at Google I had designed and launched a global company survey, supplied the logic for Google’s HR Information System (HRIS), supplied a number of important insights on employee benefits, and begun to tie these employee data sources into an Exit Risk Prediction Algorithm. While these contributions were not universally recognised, I believe I worked on the first of many of the elements of which Google’s people analytics then became.
Of course, they have taken it much further than anyone could have imagined. Out of curiosity, I recently explored my Linkedin network. Today Google’s People Analytics team appears to have 30 people working in it, and it is not difficult to find over 50 people who have contributed substantially to their People Analytics team at one time or another.
Laszlo Bock, Google’s People Operations SVP, has done a great job of promoting Google People Operations and People Analytics. Look into it – there is an article on Google People Analytics in the New York Times or Wall Street Journal a few times per year. Google has a great brand and Laszlo has been able to speak and write about what they are doing with people analytics substantially. Partly as a result of his promotion, interest in the space has grown exponentially. Ten years ago, I would have searched for a job and I would find just one role similar to what I did open across the entire United States. It really started to turn around 2012. It went from getting sporadic emails from recruiters to getting an email or call once a week about a new opportunity in people analytics. Don’t get me wrong, we have a long way to go – but we’ve gone from totally blank stares to a chance at a real conversation.
Johann: What was one of the best projects you have had so far and what were the outcomes?
Mike: I like to talk about the work I did at Google within employee benefits – partly because of how exceptional it was in terms of benefits. In many ways Google was unique. At a time most companies were pulling back on benefits – they were investing heavily. However, in other ways, Google was not so unique. Sometimes benefit programmes were chosen based on the random predilections of executives and not based on careful thought or planning. One such case was Childcare. Google decided Childcare was important, that it should be done well, that they, as a company, should take ownership for doing it well and they believed like anything else, that they could do it better than anyone else in the world. So they brought in world-renowned experts to build it, made the promise and opened it up to employees. The waiting list for the programme was almost immediately fully booked.
After building their third daycare centre, and while planning sites 4 and 5, one involving the lease and conversion of an entire elementary school, Sue decided to bring me in and have me “take a look at this waiting list problem.” So here is what I did, I:
- looked at the expected growth of the employee population,
- identified the demographic characteristics of our population,
- matched that to expected birth rate tables,
- calculated the children we could expect to have over time,
- took the children we had already,
- aged them through time,
- took the children we could expect to have and aged them too,
- and coalesced this into the expected number of children over time.
On top of that forecast, I overlaid Google’s actual cost per child, discounted by an expectation of the employee uptake on the benefit. It wouldn’t be a gross mischaracterisation to say that it looked like Google was going to the moon!
So now Sue and Google’s executives had a cost graph that they could juxtapose against their aspiration, and they could make a good decision on what they were doing relative to other investments in people. They were caught between a rock and a hard place. On the one hand, this is a great programme that many people love. On the other hand, the number of people that could benefit from it was small, with as many people “angrily waiting” on the waiting list than there were actually within the programme. There were also many more employees without children. In the end, they made a balanced decision, which gracefully distributed costs between parents and Google so that they could continue to expand the programme, but within some reason. Basically, they could use parent contributions to help fund the programme, but also reduce demand by increasing the parent contribution. Of course, some people didn’t like the solution, but it was a solution and it was data driven. This, in a nutshell, is an exercise in how all benefits should be evaluated – but typically are not. We start these things with good intentions, but nobody keeps up with the math. We take orders, we guess. Well, you will never find the optimal solution if you don’t do the math.
Johann: From my own experience, people analytics – especially in Europe – still seems to be in its infancy. Where do you see Europe compared to the US in terms of the maturity of people analytics?
Mike: I believe the US is a little further along the maturity curve now. I am not sure precisely why that is. My perception is that in some areas the US leads Europe, but in other areas the US lags.
The US happened to have a head start in people analytics, but based on the conversations I have had, I don’t believe Europe is far behind and will come on strong. Frankly, I think US business executives can be arrogant and in this arrogance miss important opportunities. We dabble in a lot of things, but don’t carry it through. For example, if you look into the contributions of Edward Deming and others in the application of mathematics to manufacturing – their ideas were not well received by manufacturers in the US at first. It wasn’t until Japan took their ideas and ran with them, to the near destruction of US manufacturing, that US business really started to pay attention.
“I believe Europe needs people analytics as much, if not more than, any other region of the world”
Will it take off in Europe? I don’t know. I think Europe has some unique challenges with employee data privacy and with unionisation and labour relations. I believe that people analytics should be used to align employees and management and I fundamentally believe it is good for both. I believe Europe needs people analytics as much, if not more than, any other region of the world. However, it is going to be dependent on entrenched powers (management and labour) to decide if this is good and to break with traditions.
Johann: Many companies are just beginning to realise the strategic importance of people analytics as an HR function and its massive benefits on both a company’s bottom line and their employees’ happiness. Where do you see the future of people analytics?
Mike: We are at a critical moment in time, a tipping point. In the history of business development, finance came along behind accounting and revolutionised the way we think about business. Then at some point marketing came and transformed sales, one of the oldest professions. Today, finance and marketing are status quo. These are taught in business schools and standard business practices. To have them doesn’t really differentiate a company anymore. It is an arms race and every move can be matched by the opponent.
“If you invest heavily in developing capability within people analytics now, you can exact an advantage in an area that is still difficult to replicate.”
People analytics is different. If you invest heavily in developing capability within people analytics now, you can exact an advantage in an area that is still difficult to replicate. Who wouldn’t want this? It is absurd to me that a CEO wouldn’t be all over this right now. Google is. It is my firm belief that organisations that don’t get this will be taken apart by companies that do. If nobody else gets this we will all work for Google some day. Maybe that won’t be so bad.
It’s hard to say exactly when we’ll reach ubiquity in HR, but hopefully within the next decade the notion that we would make HR decisions without the support of science and maths will be laughable. Do you actually want your method of making decisions to be exploited by your competition? Why on earth would you allow that?
I believe that with people analytics we have this cross-fertilisation of ideas in business, maths, science and people that put us on the cusp of a new way of understanding – a combination of human and machine learning in a single “system”. This may not be clearly understood by all, nor are most things prior to their time. However, I can see, and I believe, the possibilities to be profound. It gives me the chills to think about the things that are within our reach.
Johann: What advice or motivation would you give anyone in HR who is interested in people analytics?
Mike: If you sit in a HR department and want to introduce the idea of analytics, the best thing to do is to start with a small, very focused project. Don’t try to start by requesting a $1,000,000 budget for a two-year implementation of giant new systems and processes. It is too difficult to get all the people and data aligned.
The key to get started is to focus on a single subject, a single problem or a single question and then just go and get the data you need to address that. There is no failure in this because you remove failure as an option. If the data says your hypothesis is wrong – great, it saves you from making a real mistake. If your data tells you that your hypothesis is correct – great, you have more certainty than you did before.
Don’t try to get all the data perfect, just get the data you need.
While any company and any person can theoretically do this, I also believe in the fundamental principle of focus and value of expertise in obtaining superior results. If you are a human resources person how can you not believe in the value of focused expertise? If you are a CEO that is paid £x million per year more than anyone else for your expertise, how can you not believe in the value of expertise? If you are going to build a house would you just buy a bunch of wood and start building it? For your own safety, most civil planning institutions wouldn’t even let you. There is a reason for this. Similarly, as a professional I have to say this, “understand the professional qualifications of the person that is helping you.” This has little to nothing to do with:
- the educational degree (irrelevant),
- the brand of the company (irrelevant),
- its time in existence (irrelevant),
- or what other things the company does.
You will hear a lot of confusing things, it is worth the time to figure out who knows what they are doing and who doesn’t, and ask for help. It is that simple.