Video

Can the Occasional ‘Nudge’ Make You Better at Your Job? 

As the head of People Operations at Google, Laszlo Bock ’99 applied data analytics to human resources questions that have long been answered with hunches. His company Humu is now extending that approach for other organizations by providing AI-generated prompts to their employees. In a conversation with Yale Insights, Bock said that behavioral science, combined with an understanding of human feeling and careful attention to privacy, can help organizations run better. 


In his 2015 book Work Rules!: Insights from Inside Google That Will Transform How You Live and Lead, Laszlo Bock ’99, then Google’s senior vice president of People Operations, explained the company’s efforts to use data analytics and behavioral science to improve recruitment, human resources, and employee engagement, harnessing the power of the company’s enormous cadre of workers.

He wrote, “Our approach is to take compelling academic findings, mix in some of our own ideas, and then see what happens when you try them on thousands of people going about their daily business.”

Bock is now applying that philosophy to other organizations, large and small, through a new company, Humu, that he co-founded with fellow ex-Googlers Wayne Crosby and Jessie Wisdom. 

Humu draws on the work of Nobel Prize-winning economist Richard Thaler, whose 2008 book Nudge, written with Cass Sunstein, argues that small cues can help people make wise choices. The company’s “Nudge Engine” combines that insight with machine learning to generate carefully timed prompts, reminding employees to take actions like recognizing the accomplishments of peers or asking team members for input. 

The key, Bock told the Washington Post, is getting prompts when they are needed, rather than being reminded of best practices at a retreat or a monthly staff meeting. In the traditional workplace, he said, “you go back to your work environment, and everyone’s the same. And your behavior is not going to change because there’s nothing reinforcing it.” 

Bock emphasizes that the privacy of employees will remain paramount: for example, the employee surveys that Humu uses to set goals for its prompts are anonymized. And employees can choose to give the system access to their calendars—in order to, say, get reminders to speak up in meetings—or they can opt out. 

Yale Insights talked to Bock about what Humu can achieve for organizations and their employees—and how it avoids creeping them out. 
 

How have we been doing human resources wrong? What’s been missing from HR?

The two biggest opportunities in human resources are, number one, bringing more data and science to it. For too many years, HR has been characterized by “What does my gut tell me, what does my experience tell me?,” but there’s a lot we can learn from behavioral science and behavioral economics to make work better. The second is remembering to put a little love into the practice, because we often get focused on process and operations rather than remembering that there are actually human beings on both sides of these conversations, and caring for those individuals.

How does Humu do what you just described?

One of the things I observed in my old job running People Operations at Google was that training doesn’t always work; talking to people doesn’t always work. There are a lot of things that make it hard to get human beings to change their behavior in a way that both makes them happier and serves the organization better. 

Dick Thaler, professor at University of Chicago, coined the phrase “nudges” for small interventions that help people make better choices. At our own company now, Humu, what we focus on is using these tiny, tiny, interventions—the smallest we can possibly make—to make people happier and more productive and more innovative and more inclusive. And what we find is the smallest intervention often has the biggest result if you know what intervention to send to what person at exactly the right time.

What are you doing at Humu that you couldn’t do at Google? 

One of the lovely things about having our own company—together with my co-founders, Dr. Jessie Wisdom and Wayne Crosby—is we get to touch every kind of person in every kind of organization, ranging from companies like Sweetgreen, where 97% of the workforce are hourly workers working in a restaurant-type setting, all the way up to some of the largest financial services companies, where everybody’s got a master’s or PhD doing really esoteric financial analysis. What’s a wonderful blessing is we get to see how work is different in lots of settings and use technology and behavioral science to make work better for all of those people, regardless of where they are or what they do.

Have you been learning a lot from how people do one kind of work and applying it to another?

It’s been really interesting. When you look at how human beings perform in work across a range of job types and geographies and cultural backgrounds, some things are common. Everybody in the world would ideally like their work to be more meaningful, and that’s a universal thing. And there are ways to nudge people so they appreciate the meaning in their work or help make it more meaningful for people around them. But there are also unique differences. In some cultures, for example, people express emotion differently, and you have to be sensitive to these tiny, tiny, differences in industry and job level and in culture, because if you’re not, you might give somebody advice that causes them to actually do more harm than good.

There has been some criticism of the “nudge” approach—for instance, in a recent New York Times story. Is there a danger of employees feeling that their work is being overly monitored?

I was born in Communist Romania. I came to the U.S. as a refugee, but in Romania, the government supervised everything, to the extent where children were asked to spy on their parents, students in school were asked to spy on their classmates and their teachers. So I come from a place where the government saw and knew everything, and every company recorded every single thing about everything about their employees. That’s super-creepy, and it’s wrong. One of the premises behind Humu is that not only do we value an individual’s privacy before everything else, but that we ought not ever even violate an expectation of privacy. 

“We focus on what’s going to make you more effective in the workplace. What skills that are going to make you a better team member, make you more inclusive, make you a better manager or coach or participant?”

Companies, for example, can read their employees’ emails, and if you look at those emails and look for more emotional language you can predict if someone will stay or leave. I think that is wrong to do, because as an employee I don’t expect my boss to be reading my emails or some machine to be doing it. In a sense, we ascribe to the Hippocratic oath, but for technology. First do no harm. 

We take care of people first, and they are why our mission is making work better with machine learning, people science, and a little bit of love. We have to remember that and hold ourselves to a higher standard, and I wish every tech company would.

What is the boundary between personal and professional lives in terms of collecting data about employees?

In my mind, there is a very clean line between personal data in your home and personal life and personal data at work, and you should never cross that line—unless you’re asked by an individual to marry those data sets, and their company approves and they approve. At Humu, we don’t do that today. We gather information voluntarily from people at work, and if they don’t want to participate, they don’t have to; we’re fully GDPR compliant, so you can get your own data back. It’s your data; you own it. 

What we focus on is what’s going to make you more effective in the workplace, and in particular, what are the soft skills that are going to make you a better team member, make you more inclusive, make you a better manager or coach or participant? And all those things have to do with how you’re feeling about yourself, how the people around you are treating you, and how you perceive whether your employee cares about you as a human being or just another gear in the machine.

What should leaders do to create more effective teams?

The research we did with my team at Google, Project Aristotle, which we build on at Humu, shows that there are five things that drive team performance. It’s not team composition. It’s not your background. It’s not whether you’re a good or bad performer. The foundational thing is psychological safety. So the most important thing you can do, if you are trying to build an effective team—no matter who you are on that team—is to create an environment where every voice feels heard and everyone feels free to share their ideas. What that means is the more power you have and authority you have in that situation, the more you should relinquish it or use that power to draw people into the conversation who otherwise would not make their voices heard. That’s the key to psychological safety. That’s the key to high-performing teams.

Can managers ever be replaced by computers?

No. Think about all the managers you’ve had. You’ve had some good ones; you’ve had some awful ones. Imagine if every manager you had was as good as the very best one you’ve had. That’s what we’re building and helping with at Humu. But replace completely? It’s not going to happen, because at the end of the day, we are human beings, and that’s the same reason why a hug from your mom is always going to feel good. It feels good to be around human beings. That’s not going to go away.

CEO, Humu