Interview by Lisa Donchak
Ben Waber is recognized worldwide as an expert in people analytics, collaboration, and wearable technology. He is the leader of Sociometric Solutions, a behavioral analytics company that uses wearable sensing technology to transform how companies are managed. He is also a visiting scientist at the MIT Media Lab, where he received his PhD, and he was previously a senior researcher at Harvard Business School. Ben has been featured in Wired, CNN, The New York Times, and the Wall Street Journal, among other outlets, and his work was selected for the Harvard Business Review’s List of Breakthrough Ideas and the Technology Review’s Top 10 Emerging Technologies. His book, People Analytics, is an international bestseller and was released in 2013 by the Financial Times Press. He’ll also be speaking at WPAC2 in April!
WPAC: You’ve been working in this space for a while – tell me a bit about your background and how you got involved in People Analytics.
BW: I come from a computer science and social psychology background. I got my BA and MA doing computer vision research – using video to understand how people are behaving in the real world. When I went to MIT, I started to get into this wearables and sensors space. By chance, we’d been using next generation wearable sensor badge technology in the lab, called Sociometric badges to understand salary negotiations – how much money you might actually make based on your tone of voice, how much you interrupt someone …
These badges collect data on how people talk to each other- not content, who talks to whom, location, and movement data. They’re essentially next generation company ID badges.
We got approached by some professors at MIT Sloan who were working with some large banks in Germany. They were collecting survey data, but they thought face-to-face interaction was the most important, and they asked if we could measure this.
The early version of the Sociometric Badges looked like a tie. We brought these to this bank – we didn’t really know what we were doing, because nobody had collected this kind of data before. We said to ourselves: with surveys, we can collect data on social networks. Let’s look at that – who people talk to at work.
[At this bank], they had really good job satisfaction data, and we were able to look at it day by day, and look at what behaviors that predicted how happy people were.
At the same time, we were collecting email data. We collected survey data every day. The surveys were super inaccurate. We asked [employees] how productive they feel – those sorts of things – it was a little 15-second survey.
What was really interesting was that all the metrics that these professors from Sloan were collecting, all this email data, had some predictive power in terms of figuring out how productive and happy people were. But the Sociometric badges … just looking at a few of those features, were about six times more predictive than all of those other measures. We were predicting over 50% of the variance of those outcomes. For those scientists out there, that’s a pretty big number. Insane number.
[The bank] reorganized their entire company based on that first analysis we did. A big, multi-billion dollar company changed they way they work. And that’s what made us think we had something.
Over the next five years at MIT, we continued to develop the technology, going from company to company, not only showing that we could do better than we’d done at first, but also that we could measure the effect of that change we enacted.
A few years later, my coworkers and I graduated with PhDs, and companies came to us asking us to do this. And that’s when we founded Sociometrics.
We look at things like measuring stress, predicting turnover, and we don’t have to wait a year to publish a paper on it. We can look across dozens of technologies using our technology today … and quickly roll out and test these hypotheses. Just like Amazon does with their [A/B web design] testing, that’s what we’re allowing our customers to do with their people. I will know [in a week] week if something changed … I’ll just see it. That’s where People Analytics and HR is going to go.
WPAC: What cool research is happening in People Analytics? Anything you specifically are working on that you’re excited about?
BW: There’s a few things – short term things – I’m excited about, and long term things I’m excited about.
In the short term, I’m very excited about this idea of rapid A/B testing of management. With individual data and sensor data coming in rapidly, you can very quickly test out ideas that today are based on gut instinct. You can say “I don’t know how to lay out an office. Let’s try five different models and see how they affect key metrics. Let’s test that out.”
There are very few companies that are doing this today. Even Google, which is doing A/B testing, typically rely on surveys – not to say that’s wrong or bad, but it’s still orders of magnitude slower than what other technologies [allow]. There will be a sea-change – you’ll see people from machine learning, big analytics shops, PhDs in organizational behavior, consultants and programmers, that will subsume traditional HR today.
The next area is gonna take a little bit longer. Kent Larson over at MIT Media Lab is doing something really interesting. He’s created an office that is essentially robotic [ed. note – called CityOffice]. You have tables and chair that can move themselves around and reconfigure [the office layout]. You want to make a cubicle [setup] for 10 people, you have this little bot that will transform [the space]. If you want to have a meeting, it will rearrange into a circle. There’s a whole variety of transformations they can do. Today you sort of manually configure it.
The step beyond this is having data drive those transformations. It’s having humans in the loop for setting goals on what you want collaboration to look like. You can configure the environment in such a way that it will encourage interactions, and the space will optimize around specific behaviors. Every year, every quarter, every month, you can rearrange things. Imagine coming in tomorrow and things are just a little different.
These changes can allow you to dial in rapidly and very quickly improve performance. These slight changes are responsible for a huge amount of how people communicate at work. This is something we have a lot of research on. Our HBR article talked about data we collected, that showed that the likelihood of you talking to someone is based on how long it takes you to walk to a person’s desk.
We’re starting to see a lot of tech companies realize this intuitively. A lot of companies, based on their gut, believe a good workplace is desirable, but the next step is not believing – or proving – that it’s true, but really utilizing data and space to dramatically improve collaboration, and use People Analytics to drive how people operate at a fundamental level.
WPAC: What barriers do you see for adoption right now?
BW: We work a lot with Fortune 500 companies, with heads of global services who oversee corporate real-estate and HR. There’s a broad realization that things as they are need to change – indeed, these divisions are increasingly marginalized and not appreciated for what they do. Also, [HR organizations] are not able to adequately impact or reflect the impact they have on the business.
Overall, you see people in these situations where they oversee functions in HR, or are starting a People Analytics group – which is happening at a lot of companies, and I’m sure we’ll see them at PAC – traditionally, these divisions of large companies don’t have a budget set aside for People Analytics.
You won’t see heads of HR or People Analytics say data doesn’t matter at all – they do [say that], even if they haven’t been to PAC or one of these events. The difference is, when it comes to building a team, investing in the tech to make this happen … [People Analytics] hasn’t existed before, and you don’t have that kind of money set aside.
Typically, the way these [HR] organizations are viewed – they’re viewed as a cost center. The [HR] orgs that are gonna succeed, they’re the ones where the CEO says, “this is a strategic priority for us.” They’re gonna hire people with these capabilities, buy technologies like sensors and email analytics. These are the companies that are forward thinking and are gonna succeed. You need to get buy-in from the rest of the C-suite to make it happen.
WPAC: What advice do you have for those interested in implementing People Analytics solutions?
BW: It’s a good question. Really, you do need buy-in from leadership in the company.
If you run a division yourself, maybe you can bring in a few people in-house to build a People Analytics capability. Really, to get serious about it, you need to hire a whole bunch of people to make it happen, and bring on a lot of technology to make it happen. Relying on surveys is a stopgap solution. When you’re presenting to a CEO and you want to show an ROI of what you’ve done, things like surveys are a lot harder to sell. Things like email analytics, sensor data – you can really put hard numbers on that in terms of improving performance, improving retention.
There’s a lot around analysis of resumes to staff people better. That is important. Long term, though, it’s not about bringing in the most brilliant people in the world. If we hire a pretty good group of people, get them to collaborate effectively and share ideas, that’s where the benefit of People Analytics really comes in. That’s what companies need to emphasize, and what people need to emphasize to their leadership.
If you invest today in hiring people with these capabilities, people from machine learning, big analytics shops, PhDs in organizational behavior, consultants and programmers, it’s a big investment. But in ten years, everyone is going to do this.
Twenty years ago, Amazon was doing A/B testing on their website. Everyone thought it was crazy and a lot of effort.
But we’re seeing greater benefits in the workplace [from People Analytics solutions] – the smallest improvement our customers have seen is a 7% increase in sales. If you look at other companies that are doing email analysis, you’re still seeing a big lift you can’t ignore.
In 10 years, everyone’s will have to do this. But if you wait until then, organizations will be changing day-by-day based on their analytics. In the years you’re catching up, your competitor is growing 20-30% more per year.
You need to do it in the next couple of years, otherwise you’re going to be left behind.
WPAC: Anything else you want to share?
BW: [People Analytics] is hard to do today, [and] that’s why you need to do it. Which is weird, because typically you want to wait to do things like this. But it’s similar to marketing. People in traditional HR departments need to think about how it’s structured moving forward. You can’t get by going on gut instinct anymore. You maybe have a few years of that left. You can see leaders [in this space] and companies who are doing poorly, and the middle of the pack will come in the next few years.
Okay, so Google’s doing it, and they’re weird. When a big, French company does it – the point is, even those kind of companies are doing this. And that’s very validating.
To learn more about Dr. Waber’s work, connect with him below:
MIT Media Lab: http://web.media.mit.edu/~bwaber/