Humu is using machine learning to create a happier workplace

 

November 20, 2018

 

Wayne Crosby, co-founder and head of product & engineering at Humu, visited the Sutardja Center’s Silicon Valley Innovation Leadership Week executive program to talk about Humu, a startup using artificial intelligence to help make employees happier at work.
Wayne Crosby, co-founder and head of product & engineering at Humu, visited the Sutardja Center’s Silicon Valley Innovation Leadership Week executive program to talk about Humu, a startup using artificial intelligence to help make employees happier at work.

Wayne Crosby, co-founder and head of product & engineering at Humu, visited the Sutardja Center’s Silicon Valley Innovation Leadership Week executive program to talk about the next industry being revolutionized by artificial intelligence and machine learning: human resources.

In Crosby’s view, the HR industry (which he estimates to be about a $400 billion market worldwide) has hitherto been the domain of intuition, without much technical analysis into what makes employees happy and successful. But considering that most people spend a large part of their lives at work, and with today’s ability to run experiments and use data to gain insights, he believes that employees can be happier and more productive, with a bit of nudging.

According to Crosby, research shows that performance and retention are driven mostly by employee happiness, and the main building blocks of happiness at work are meaning, trust, and empowerment. Meaning comes when an employee feels that their work is adding value for another human being. Trust means that workers feel psychological safety at an organization, and are comfortable being themselves. And empowerment means that employees have broad control in how their own work gets done.

“The science is really clear that these three elements are able to drive happiness,” said Crosby.

So with an idea of what makes employees happy at work, Humu can run large experiments involving small interventions with employees to see if the nudges actually can change behavior and make employees happier.

Crosby mentioned an example at Google to show how a large experiment like this would look. HR managers noticed that it was taking nine months on average for new hires to be on-boarded. After interviewing managers to ask how employees could be ramped up more quickly, managers had very similar recommendations: 1. Make sure the employee has a working laptop on day one, 2. Ensure the employee has a good startup project, 3. Ensure managers have a one-on-one scheduled, and 4. Make sure employees know where to get help. To make sure these basic things happened, HR simply sent two emails to new employees with this information. These emails were so effective that they reduced on-boarding time from nine months to six months. Through experimentation it was verified that these emails were actually the causes of the shortened on-boarding time, saving Google approximately $400 million in productivity.

Crosby first became interested in making the lives of workers better when he co-founded Zenter Inc. in 2007, while juggling time with his first newborn.

“The night I applied to Y Combinator, my wife and I found out that we were expecting our first child,” said Crosby. “That was a scary moment for trying to start something new and innovative. That shaped how I started to think about how to use data and information to actually make the workplace better.”

Zenter Inc. received funding from Y Combinator and very quickly met its goal of building next-generation presentation software. Zenter was acquired by Google Inc., and not long after became known as Google Slides.

In 2009, executives at Google, where Crosby then worked as VP of engineering after the acquisition of Zenter, were inspired by a story about NBA basketball player Shane Battier in the New York Times. Battier was an interesting case because while he didn’t have impressive statistics in the traditional metrics tracked by the NBA (e.g. field goals, rebounds, steals), it still seemed to be the case that his teams performed better and his opponents’ teams performed worse when playing against him. In order to discover how Battier was making it work, analysts had to dive deep into the data to understand what he was doing that was working so well. For Google, it made them wonder:  Who are the Shane Battiers of our organization? Are there employees that are making outsized contributions that are not being acknowledged?

This insight led to Google HR executives identifying Crosby as a great manager, a Shane Battier for Google. For this, Crosby received an award, but perhaps more fortuitously it led him to meet his future co-founder, Laszlo Bock, who at the time was senior vice president of people operations. Bock and Crosby immediately hit it off over their curiosity for how teams work and what makes them perform well.

After this encounter, and watching a talk by Google Fellow Jeff Dean showing how a machine learning algorithm learned to detect cats on YouTube, Crosby began to think in earnest about how data may be used in new and interesting ways.

Besides becoming more interested in the power of machine learning, around this time Crosby also learned about research by Jessie Wisdom, a behavioral economist working to understand why people make the food decisions that they make.  Wisdom is an expert in Nudge Theory, a concept popularized by Richard Thaler, who won the Nobel Prize for his work on the subject.

Crosby describes a nudge as “a stimulus that you add to the environment, that does not remove choice, that has a material impact on behavior that can be measured.”

One very popular example of how a nudge works in practice is with organ donation. Behavioral economists have found a very simple explanation for the variance between people who do (or do not) donate their organs upon passing: whether or not the form (e.g. when one renews their driver’s license) has the organ donation checkbox checked by default or not. When the donation checkbox is checked by default, organ donation rates are much higher, as in Spain which has one of the highest rates of organ donation in the world.

Crosby saw the power of this firsthand at Google when a small nudge was implemented in the cafeteria by Wisdom’s team. To help Googlers make better food decisions, and to ward off the weight gain that was common among new hires thanks to abundant snacks, they implemented a very small nudge intervention. Where self-serve snack dispensers that included healthy options, such as peanuts or dried fruit, were placed side-by-side with candy, the candy dispensers were made opaque so that the colorful rainbow of the candies could not be seen. This worked quite well: they found the New York office consumed 800 fewer pounds of M&Ms.

The power of the nudge was the final piece of the idea for Humu. Bock, Crosby, and Wisdom co-founded the company together to see if they could use their combined experience with data analytics, human resources, and behavioral economics to make the lives of employees better.

“When we talk about happiness,” said Crosby, “what we are talking about is a feeling of contentment, of belonging, of waking up and not dreading going to work, but actually being excited to be there, feeling like you are contributing to something that is bigger than what you can build with your own two hands.”