What can we learn from failure?

Google’s Banks Baker, Chicago Booth’s Harry L. Davis, David Hill of TIL Gaming, and Eunhee Sumner of Starbucks discuss the insights we can gain when we don’t succeed

Nov 21, 2018

Sections Strategy Video


Harry L. Davis
Roger L. and Rachel M. Goetz Distinguished Service Professor of Creative Management, Chicago Booth

Banks Baker
Head of global product partnerships–search content, Google

Eunhee Sumner
Vice president, continuous improvement–customer, product and marketing organization, Starbucks

David Hill
CEO, TIL Gaming

Tell us about an experience when you failed professionally, and what you learned from it.

Baker: I started my career in Silicon Valley in the mid-1990s, with a fairly high-flying dot-com. A small group of us broke off to do a start-up. It ended miserably in the dot-com bust. It was a failing, not just of the market but of our business model. It was a failing of our skills. I learned two things. First, because I was inexperienced, I hadn’t understood the risk well, and secondly, although I blamed myself a lot for quite a while, I learned over time that I needed to spend more time analyzing the actions and the decisions and less time judging myself.

Sumner: My biggest failure was being on the front lines when Washington Mutual failed. I led the mergers and acquisition team. We raised about $7 billion in funding from private-equity firms in May and June 2008. Then, on September 25, we were seized by the Federal Deposit Insurance Corporation and the bank was sold to JPMorgan Chase. 

There are a lot of lessons. One, what were the risks that we were taking and why were we overexposed? Another is: How did we manage our business, in terms of operations and making sure that we had visibility and risk-mitigation tools in place? Also, leadership: Were the right leaders in place to get us through this change? Were we being quick enough to make those decisions and change out executives? When you can’t make those tough decisions fast enough, it really affects the way that you’re running your business. 

Hill: Years ago, I was a professional gambler. I did statistical arbitrage on sports betting in Las Vegas. Luckily, the stakes were fairly small when I started, because I literally lost everything a couple times. One time, I sold my beloved motorcycle and we restarted. The second time, my business partner pulled all of his money out of his 401(k). Early on, a lot of my failures were really in risk management, and not understanding that having an edge didn’t really matter. Since then, I’ve learned about risk analysis.

The biggest failures I’ve had as an entrepreneur are more on the personnel side. When I’ve been enamored with people and brought them onboard, the biggest mistake I made was letting them stay there too long when things didn’t work out. There was one manager whom I very much liked personally but there were things going on in his life [that negatively affected his work, which I turned a blind eye to]. As a result, I wasn’t doing a good job and I started losing other staff. They stopped respecting me and the company because I didn’t make the move [to fire him] quickly enough. Over time, I’ve learned to pull the trigger earlier, be a little bit more cognizant of what’s going on with my staff and how any one poison person within the staff can really destroy the morale.

Davis: I came to Chicago Booth, not being an economist and not being particularly quantitative. I was teaching marketing at the time, and I thought I had to be really very rigorous. I gave a really tough multiple-choice exam. I felt very proud about giving something that was really tough. There was one student who was obviously engaged and really smart. I ran into him and asked, “What did you think of that exam?” He said, “I expected more of you.” It was like somebody had stabbed me. It was an incredible learning experience because I realized this had more to do with me. Why was I doing this? I should have just been doing the kinds of things that were more who I was, rather than trying to pretend that I was an analytic, mathematical person. 

What are the mechanics of learning from failure? 

Davis: I see failure as an unexpected outcome. Outcomes can sometimes be better than we expected and sometimes worse. In some ways, both are failures. If something turns out to be much better, why was our forecast so flawed? We ought to look at that as much as we look at the other side of the ledger. Every day we have opportunities to experiment, to try things, and to learn. I’m trying to get people to think about the workplace as an environment for ongoing experimentation. What makes an experiment really valuable is often something that doesn’t work out as we thought. 

Hill: It’s business as a series of hypothesis testing. Take the time to really test an idea. When you test multiple ideas simultaneously, attribution becomes impossible, making it difficult to learn from the experience. I was working a lot of hours when I first took over my business, 80–100 hours a week. When I cut back to 22–25 hours, I apologized to my regional manager. He said, “Are you kidding? This is the best thing ever. You were doing everyone’s job. Nobody knew what to do because you were the only one making decisions. Now, you spend that little time you have with us, the managers, and we’re able to actually manage the way you want us to.” That was my failure in thinking that I had to do everything. I was doing nothing to make myself replicable.

Baker: As leaders, we have to create a culture that is a safe place to experiment and to push boundaries. We do that by looking at the collective knowledge that comes from analyzing failure, to share best practices and bring the overall group higher and move with greater velocity. It’s important for leaders to instill that. We have to be proactive.

Sumner: Working in the VC business, I learned a critical lesson that has been helpful throughout my career: make the call when you see issues and red flags. Don’t be a bystander and wait for things to unravel. Get it back on the rails as quickly as possible.  

Davis: There was a neurobiology professor at the University of Edinburgh who talked about having two academic résumés. A regular résumé mentions all the awards we’ve received, all the papers we’ve had published, and all the grants that we’ve secured. She has another résumé, three times as long, of all the things that she requested that she got turned down for. She uses this in working with young scientists to say, “Which résumé is going to be more meaningful for your life?” You’ve got to keep at it. You’ve got to be persistent. You’re going to be told “no” many times. It’s not a failure. It’s another reason to keep going. 

Many organizations have high turnover of personnel and poor institutional memory. How do leaders institutionalize the learnings from past failures, so they don’t just keep making the same mistakes?

Baker: A lot of organizations do postmortems if something’s failed, to look at the causes. At Google, we sometimes do a premortem. We get the stakeholders in a room and role-play: “It’s a year from now and our project has failed miserably. What was the reason?” You start to brainstorm and you take advantage of that collective knowledge because, again, you might not have continuity of leadership. You might not have continuity of a team lead or an engineering lead or a product lead but, with the collective knowledge of all the stakeholders, you crowdsource your ability to mitigate risk. I don’t know that we do that quite enough. I haven’t seen it in other businesses. But when we’ve done it, I’ve found it really valuable as a way to try to uncover those obvious mistakes because, frequently, when something fails, it’s not a massive market-driven issue. In some cases, it’s fairly obvious in hindsight. It was a failure of design. Or someone forgot to flip a switch or something. Doing the premortem leads to more success, or at least it mitigates more risk, because you can start to really collectively think about your own failure and the causality of that. I think that’s very healthy, as opposed to just individually assuming that you have it right and marching forward. You don’t want it to pause the project too much, but I think it’s an important gut check as you head in.

Sumner: You’re talking about building a culture. Failure isn’t a bad word, and it has to be just as important as asking, “What did we learn from our successes?” Celebrating that more, and being comfortable with talking about it openly, is important. You can’t plan for everything, because there’s going to be fallout that you didn’t think through or didn’t catch. It’s like a wedding. There’s always going to be something that goes wrong. You plan for as much as you can, but you’re building that muscle for the team to start figuring out: How do you work better together? How do you increase communication? How do you think things through so you can do the problem-solving on the fly? 

Baker: Thinking of failure as an event is not that useful. Recognizing the event of failure in a larger ecosystem of learning, and being able to pivot and move, and having the cultural safety to be able to operate in that environment is super important.

Davis: I think it’s not irrelevant to think about an organization as being an “experimenting organization.” Experimenting is really part of the core competency. We experiment. We’re constantly experimenting. Hopefully, we’re experimenting in ways that are going to avoid large downsides. The more we do incremental experiments along the way, the more likely we are to avoid large disasters. 

Hill: There’s a man in my company who’s been in the gaming industry for 35 years and brags about how he drove the previous casino he ran into complete bankruptcy so that they shut the doors. He says, “I was that guy that couldn’t change.” He helps me realize failures that he’s had, and it saves me from myself. What impressed me so much—when I interview candidates for senior positions, they tend to talk about their accomplishments—is the first thing he talked to me about when I interviewed him was his failure and what he had learned from it. That’s why he got the job. He’s my right-hand guy and he’s amazing. 

Sumner: For Starbucks, you have to figure out how to stay current, how to be innovative, how to fail fast, and how to build this whole agile methodology. Before, it used to be these big projects, big teams: build it. And huge infrastructures, huge investment. It feels much more catastrophic if that huge project fails. As we become more nimble, we need to become quicker to fail, to learn from it, and to recoil and go back to the next phase to keep doing those incremental moves. The pace of business these days is so fast that you can’t wait for these huge launches. 

Baker: Launch and iterate, launch and iterate. One of the challenges as you get larger and larger is: How do you keep that culture? How do you keep that when you go from where Google was to where Google is, or where Starbucks was to where Starbucks is? Keeping that culture and keeping that speed is a real leadership challenge.

Hill: One of the biggest keys to learning from failure is owning it. One of the things that I see that really concerns me is when someone fails and they say, “This happened and I got unlucky”—everything’s external. Most of that was usually foreseeable, at least as a potential. If you own your failures and look at what you’ve done, that’s where you learn.