Eliminating Friction at OpenAI

Fidji Simo, CEO of Applications, recently said, “Companies rarely become bureaucratic all of a sudden. It happens one unnecessary meeting, one extra approval, one small frustration at a time. Each makes it only 1 or 2% harder to do your job, which is why they’re easy to ignore, but those frictions compound.” She offered this quote in the context of an article about OpenAI's process for identifying "friction" within the company and cutting through the bureaucracy to eliminate these pain points. Fortune's Emily Forlini explains:

At OpenAI, a special process—and a magic word—helps employees instantly cut through layers of bureaucracy and clear bottlenecks slowing things down.  It’s called “friction.” And it has the power to supersede just about anything a team is doing, often carrying the force of CEO Sam Altman or President Greg Brockman with it.  The process starts when someone emails a special friction@ address to report an internal bottleneck. The issues range from a technical system that isn’t working, a process that’s not having the intended result, or office-related frustrations, such as there not being enough IT vending machines. OpenAI’s leadership team triages the emails sent to friction@, moving forward with ones they deem worthy. To help alleviate full parking lots, for example, the company started a pilot to prioritize spots for those with long commutes. Another time, employees asked for a better process to grant API credits reliably and at scale. If the matter is considered important enough, Altman or Brockman will get involved to ensure it’s resolved.

A former OpenAI employee who spoke to Fortune praised the friction system as an effective way for leadership to get ground-level feedback. The process helps teams keep moving forward, tackling issues before they can fester, the person said. That’s especially important as OpenAI’s headcount grows to more than 8,000 employees expected by the end of this year, with offices across the U.S. and the globe. “It’s a good thing—literally anybody can complain about big company bullsh*t directly to Sam and they can act on it,” the former employee said. “If you want to ruthlessly cut through bureaucracy, you have to be ruthless about it.”

The process reminded me of the fantastic book, The Friction Project: How Smart Leaders Make the Right Things Easier and the Wrong Things Harder, by Stanford Professors Bob Sutton and Huggy Rao.  Perhaps Altman and Simo read the book and built this process in light of those insights about eliminating friction.  One word of caution though - Forlini writes in the article that some employees have complained about the tremendous burden and pressure it puts on people to drop everything and address an issue that has suddenly been escalated to Altman and Simo.  As always, balance and moderation is critical.  You want a process for cutting through bureaucracy, but you don't want it to become abused by those annoyed at small inconveniences, or by those trying to reallocate resources to their pet projects or ideas.  

Career Advice: Run to the Mess

Several years ago, former Market Leader and Likewise CEO Ian Morris spoke to my students at Bryant University (where Ian graduated with a Bachelor's degree in the early 1990s before completing his MBA at HBS).  Ian offered a few wonderful pieces of career advice.  First, he always told the students to "leave well" when changing jobs.  Give extra notice, help train your successor, and preserve relationships.  You never know when you will cross paths with those people again. Second, Ian suggested that they "run to the mess" if they want to build new skills and accelerate their career progression. Ian mentioned that many people avoid difficult situations because they fear failure. Yet, he described these tough circumstances as the best opportunities to truly shine, to make your mark.

A recent Stanford interview with Jane Fraser, CEO of Citibank, reminded me of Ian's advice. Fraser also commented on the value of taking on the tough jobs during her career, even if they were not the ones that others found attractive and interesting. Fraser said,

Interviewer:  Why did you pick these tough jobs?

Jane Fraser: Yeah. I mean, because someone once asked, “Is it because none of the men wanted to take them?” There might be a bit of truth in some of that. I think one of the things, I’m not a tweaker so I never wanted a job where you’d just be tweaking things to make it better. I think I’d be awful at it. I think a lot of people in this room would be awful at tweaking jobs too. It’s more fun to be bold in making changes. And then if something’s not in great shape, it’s amazing how you get left alone to fix it. Whereas if something is actually going well, you have everybody giving you helpful advice and opinions on what you can be doing better. So one of the fun things about turnarounds or changing things more radically, quite often you’re left a little bit more on your own.

Then when it starts getting successful, then everyone jumps onto the train and heads that way. But I’ve always quite enjoyed transforming things a bit more and then you learn a playbook of what you need to do and you can repeat a lot of it in different situations like get the team in place that you need as a first step because otherwise you’re going to be doing everybody else’s job and be really lousy at the one you’ve actually been put in to do. So there’s just different steps that I put in place each time now and they’ve been helpful so far.

Why We Love It When Customer Service Professionals Remember Us & What We Like

This summer, my wife and I have traveled quite a bit.  We have had the privilege of some amazing customer service experiences.  One thing, in particular, has stood out.  Some customer service professionals got to know us, remembered our names and our preferences, and anticipated our needs and desires before we articulated them.  Sometimes, they remembered our favorite drink order.  In other cases, they asked us how we enjoyed our previous day's adventure.  We loved it.  Why?  Well, of course, humans love to feel valued.  We enjoy being seen and heard.  These professionals clearly had listened to us, gotten to know us, and had made the effort to make a connection with us.  In turn, we learned their names, and we got to know them.  When our breakfast waiter at the hotel had a day off, we asked him about the day when he returned.  He appreciated the chance to share what he had done.  These professionals were not in a hurry to simply fulfill our order.  They took that extra time to interact with us.  They made a human connection.   

Sometimes, these professionals made mistakes. Honestly, we didn't care much at all. We were very forgiving because we felt valued, and they were clearly trying to help us.  They apologized. We let it go.  Companies cannot prevent all mistakes.  However, they can create a customer experience in which most people will be quite forgiving.  

The lesson for companies is clear: Don't just value efficiency in customer service.  Make sure that you emphasize the value of building human connection.  Encourage your people to listen actively, learn customers' names, and get to know them.  Don't penalize them for taking an extra minute to fulfill an order.  Weigh the long run benefits against that tiny drop in "productivity" that you might experience.  Moreover, companies should consider helping their employees build their active listening and empathy skills.  Some people are naturals at making human connections. Others will need some training.  Invest in building those capabilities; the payoff will be great. 

Interpersonal Coordination on High Performance Teams

What's the key ingredient to high performance when teams function in demanding, high-stress environments? Mary Waller and her co-authors have studied this question in the context of airplane cockpit crews. They published a paper recently titled "More than a feeling: rapport and complex task performance in teams."  They examined three elements of rapport within these teams while they worked in flight simulators:

1. Interpersonal liking: how positively did the individuals engage with one another
2. Emotional positivity: how optimistic and positive was the emotional tone of their conversations
3. Interpersonal coordination:  How well synchronized was their verbal communication and action

Sally Blount summarized their key finding: “Positivity and liking are great. But unless you can translate that positivity and liking into improved collaboration, it doesn’t have a significant impact on team performance.”  After analyzing video recordings of the teams in the simulator, the researchers observed how the coordination among the high-performing teams tended to be more complex, nuanced, and synergistic.  Kellogg Insight summarized the findings:

Overall, the researchers found that, on average, the teams with the highest levels of interpersonal coordination communicated more often, especially with commands, observations, suggestions, and laughter. The interactions between these pilots also were “significantly more complex” than the interactions between pilots with lower interpersonal coordination.

“In the low-interpersonal-coordination groups, you had less communication and also less encouragement and building off of each other,” Blount says. “So, if things aren’t going well, you might be saying, ‘This guy isn’t helping me, so I’m going to hunker down,’ versus ‘Okay, we got this,’ and ‘I see that. What do you see?’”

Not surprisingly, the pilots themselves did not always perceive the link between rapport and performance correctly.  

Interestingly, when liking and positivity were high, the pilots themselves thought they performed better, but that perception did not always translate into better expert evaluations. The tendency to associate more-positive interactions with better performance is common within teams, but it can be misleading and is a well-known contributor to groupthink, Blount says.

Constraints Are Good For Us

Do constraints and limits enhance creativity and innovation?  Yes, David Epstein argues that they can, bringing together several strands of research, coupled with some amazing stories, in his terrific book titled Inside the Box: How Constraints Make Us Better.  

For example, Epstein tells the story of the terrific young pianist, Keith Jarrett, who was scheduled to play a concert in Cologne, Germany.  Unfortunately, the organizers brought the wrong piano to the venue, and it was woefully out of tune.  A rushed effort to remedy the situation provided only a partial fix.  Forced to improvise with a far-from-optimal piano, Jarrett produced an incredible concert performance.  Epstein writes, "That recording, of the undersized, partly tuned, sticky-pedaled piano, eventually became the bestselling solo piano album of all time."

Epstein grounds his argument in solid research studies, as well as terrific stories. For example, he describes the work of Catrinel Tromp.  She wrote an article titled "The Green Eggs and Ham Hypothesis: How Constraints Facilitate Creativity."  Tromp asked research subjects to develop two-line rhymes for greeting cards.  She found that her subjects were more creative when she imposed significant constraints on them.  She likened this effect to Dr. Seuss writing his famous story after making a bet that he could write an entire children's book using just fifty words.  

I highly recommend the book.  It is an easy read, and his arguments are compelling.  He shows what humans can do when faced with limits.  They improvise, innovate, and create amazing things under certain conditions.  

Staying Grounded: Find the Right Sounding Board

Former Baxter International CEO and current Kellogg Professor Harry Kraemer has written a terrific column titled, "What Every New CEO Should Do In Their First 30 Days."  His lessons apply to people who take on new leadership roles at all levels. My favorite piece of advice focuses on finding the right sounding board.  You need to keep a few truth-tellers close to you.  These are the people who will give you the unvarnished perspective.  They will help you stay grounded too.  Kraemer writes,

The higher you rise, the more essential it becomes to have people who will tell you the truth. For me, that person is my wife, Julie. Whenever I was promoted—including when I became CEO of Baxter International, a $12 billion healthcare company—she would always say how proud she was of me. In the next breath she would remind me, “Harry, we’re not going to change the way we live, right?”

In the same way, you need a sounding board composed of a variety of people, such as a spouse or partner or close friend, a board member (or two), a former colleague now leading another company, or a professional coach. Each of these people has a unique perspective on you, your leadership, and the kind of support and honesty to remind you of who you are, as a person, not just as a leader.

Are You Listening to Me? How Can I Tell?

Are you listening to me? This thought comes to mind for many of us during conversations. How can we tell that someone is listening? Conventional wisdom is that body language and non-verbal cues indicate attentiveness. Is the counterparty nodding in agreement, making good eye contact, and leaning forward into the conversation? If so, they must care about what I'm saying. If not, then they are clearly disengaged.

New research suggests that this conventional wisdom may not be right though. In an article from UCLA's Anderson Review, Carla Fried reviews a stream of research by Hanne Collins and several co-authors. This research finds that people systematically overestimate the extent to which others are listening closely to them. They overestimate because they misread utterances or nonverbal cues... they think that the quiet "mm-hmm" or the nod of agreement signal close listening, when they may not represent that at all. Listeners may be using these conversational devices to "fake" close listening.

Collins and her co-authors instead suggest "conversational listening" as the remedy. Much like a stream of other work, including a great book by Charles Duhigg, they suggest that listeners paraphrase what they have heard and ask for confirmation, ask clarifying questions, and refer back to points made earlier in the conversation as ways to engage more effectively in a conversation.  In short, you must speak, not remain silent, to listen effectively.   Of course, you can not speak too much.  Interrupting without the intention of understanding more effectively can be highly counterproductive. Collins and her co-authors write, "“Conversational listening is a key building block of human social functioning. Information transmission, interpersonal connection, conflict management, happiness — the key foundations of human flourishing — hinge critically on our ability to hear, understand and respond to others.”