Eric Ries is an entrepreneur, the bestselling author of The Startup Way and The Lean Startup, and the creator of the Lean Startup methodology, a global movement in business practiced by individuals and companies around the world. Ranked one of the world’s most influential management thinkers by Thinkers 50, Morten Hansen is a management professor at University of California, Berkeley and the bestselling author of Collaboration and Great at Work. The two recently sat down to discuss why every business should be willing to experiment, and even fail, in order to foster innovation and grow in a positive direction.
This conversation has been excerpted from the “Open Minds” video series on American Express OPEN Forum®. View the complete video and series for more on management.
Morten: What are some of the key structural changes that businesses need to implement to move to new modern management?
Eric: It’s important to remember that we’re talking about entrepreneurial management as a complement to general management, not as a replacement. There are certain problems that businesses of all sizes face where the logical structure is [to have] internal startups.
A startup is really different from a committee or a project team that people are used to. First of all, it’s a full time job for a small number of people. Amazon famously calls these two-pizza teams, a team no larger than you can feed with two pizzas. It’s a small but dedicated team that’s completely focused on solving one problem. It’s funded in a different way than most corporate projects are funded, with a fixed or metered budget that says you only have 30 days, 90 days, and it’s yours so you really can spend those limited resources however you like. But in order to unlock more funding, you have to accomplish specific milestones.
That scarcity mentality really focuses the team, that scrappiness that we associate with startups, on the accomplishment of a customer-facing goal.
Morten: It’s not about giving that internal startup lots of money upfront.
Eric: No, the more funding, the worse. We all know famous stories about startups that have been overfunded, but corporate startups are 100 times worse. They tend to be massively overfunded at the beginning. It’s very important to start small or you never get that scarcity mindset going.
Morten: How do you get people inside of a company to work on a scrappy startup? You’re not going to get the money, you’re not going to get the resources, it’s going to be scrappy. Who applies for that kind of position?
Eric: I was once giving a talk on this topic to up-and-coming finance executives at a big multinational. I was explaining that if they wanted to become part of the solution instead of being part of the problem, finance has a very important role to play in setting alternative KPI’s, different compensation mechanisms for startups, in particular, the way you should handle compensation, for example, if you’re working on an internal startup, [is to not] get the corporate bonus plan [and] take a reduced salary. You don’t have the same upfront compensation as other people in the company, but you get equity. You get a portion of the profits that that startup ultimately generates for the company.
In the event, people were almost throwing tomatoes, hissing and booing. It was really very negative. [But] as I was leaving the room, several people came up to me, and [asked if] the reduced salary was available. “Where do I sign up for that? Is that possible?” For a certain kind of person, that’s a trade they’ll make because they want to have disproportionate impact, and they don’t need the money that they’re being paid right now. They would much rather gamble that on a lower probability of making a lot more money down the road.
Morten: But the incentive system or compensation system in an established company is set in a certain way. And what you just articulated is completely different. That’s one of the things you need to change.
Eric: It’s really hard. I’ll give you another example. A manufacturing company team had come to a three-day workshop about lean startups that the company had sponsored them to do. It was an American company, but they were going to sell this product into a developing economy, in a market where they were not familiar.
In the course of the conversation, the traditional plan they had come to the workshop with was to spend three or four years designing this technology, making it perfect, launching it globally all over the world. We started to have a conversation in the room about what would be required to get this product to market sooner. One of the engineers said, traditionally, we would design and assemble all the components linearly, one after the other in sequence, but there’s a technique called set base concurrent engineering, which is famous in manufacturing circles for getting cycle times down. We could design and develop the components in parallel and then bring them together at the end. As a result, we could get to market a lot sooner.
Everyone in the room was getting excited [with] this much better plan. They could get the minimum viable product into market sooner [and] have a much greater chance of success. I said, “Are we going to propose this plan to senior management and do it?” and they looked at me like I was crazy.
If everyone in the room agrees this is a better plan, why not do it? They said “You don’t understand, our personal metrics of success as functional engineers in this project would be destroyed by this project. Because we do things in parallel, there’s going to be more re-work, there’s going to be more possibilities for mistakes, they’re going to have to be corrected.”
Morten: It’s the wrong metrics.
Eric: If they’re the wrong metrics, not only do they make it difficult to innovate, they make it impossible. You’re actually paying people to prevent innovation. When we put it that way to the company senior managers, they were eventually going to change, but it was quite a difficult process.
Morten: There’s a saying, “What gets measured is what gets done and focused on.” That also brings up this other key change that is necessary, and that is tolerating failure. There’s lots of academic research that shows in big, established companies, you’re not supposed to fail. I know you write about that in the book, and you have a really fun story about that.
Eric: I was in a very established, big company, meeting with one of the process leaders for the manufacturing division. On his desk, he has a mug that’s distracting me [that] says “Failure is not an option,” printed on it. I was thinking, “What kind of life would you have to lead if failure is not an option?” I was trying to think [about] the entrepreneurs I know and admire [and] what their mug would say.
We’d be lucky if we only failed 12 times a day. Their mugs would say “I eat failure for breakfast.” There’s this totally different mindset around failure required. Now, here’s the key idea. The reason he could have a mug on his desk that says “Failure is not an option” is because in certain kinds of work, there’s a presupposition that if you fail, it’s because you didn’t adequately plan. In his world, if you’re diligent with your planning, if you’re disciplined in your execution, you can guarantee that you’ll have no failures, barring an act of God.
In an entrepreneurial context, no matter how well we plan, no matter how disciplined we are in our thinking, failure is inevitable because of the incredible uncertainty of the underlying thing we’re trying to do. So it’s really important in a corporate context, we can’t tell people we’re going to normalize failure and make it okay, because some failure is borne of incompetence, and that’s still bad. But if we’re so afraid to fail, we’ll never be able to take any chances.
Morten: You talk about this concept of productive failure versus any failure. What’s the difference?
Eric: I actually recommend that companies put this on their performance matrix for employee performance reviews and say, “How many productive failures did you have this year?” That immediately changes the conversation because anyone who claims they had zero failures in a year is either lying or were so conservative about what they undertook that year that there was no possibility of any innovation.
A productive failure is not a failure that you were incompetent or didn’t know what you were doing. It’s a failure that led to a success. So it led to a pivot, a change in strategy without a change in vision. [One that] allowed something new to unfold that wasn’t possible, or you have a kind of failure where you realize inexpensively, we shouldn’t even be doing this project at all.
I was working with an American company that had this new healthcare product that they were excited about, and they were going to invest a million dollars in developing this product. The team said, “Before you give us a million dollars, give us $10,000 to do an experiment.” So they did a $10,000 experiment, which looked [slightly] positive, and asked for $100,000 to do the next experiment. $100,000 in, they realized that there was no market for this product [as] they [had] thought, and their leap of faith assumptions were wrong. They came back to management and recommended they cancel this project.
In that company, that had never happened before. The idea of an employee declining extra funding was crazy. By containing the total liability of these initial experiments, they were able to fail very inexpensively. The challenge in that case was to get management to see that not as a $100,000 failure, but as a $900,000 savings.
Morten: I think it’s a great example, and sometimes, in companies, it’s the more money you have and more resources you have, the more powerful you are, but it’s just the opposite.
Eric: That’s the lesson of startups. Sometimes small things become huge.
Morten: That’s right. What I found in my research is that if you want to learn, if you want to progress, you have to conduct small experiments, change the things that you do in your job, like a salesperson who’s tweaking the sales pitch to customers, and be prepared to fail. Experimentation at work involves failing and you cannot learn if you don’t allow for that. We found [that] some people were willing to experiment with the sales pitch. Others were never going to change what they were doing because it was good enough. But good enough is not the same as being great.
They didn’t have what I call a “learning loop.” They were not trying to experiment, get some feedback from it, modify their approach and keep on doing things. They just kept on doing the same thing over and over and over again, and they were stuck.
Eric: I think that requires us to redefine what failure means. Because the interim things that happen when you’re doing a learning loop, when you’re doing that experimentation, those aren’t failures. Doing the same thing over and over again, and getting the same mediocre result, that’s failure.
Morten: Exactly. We have the wrong definition of failure. There’s another thing in your book that really struck me, and it’s the concept of innovation accounting. Tell us more [about] that.
Eric: The purpose of accounting is not just about allocating money and where the money went. It’s an accountability system for figuring out which teams are doing a good job and which ones aren’t.
In most organizations, we just look at ROI. We see which teams have the best return on investment and give more money to them. But if you really believe that innovation is going to follow that hockey stick shaped pattern that we see with all famous startups, what’s the ROI of the project during the flat part of the hockey stick? It’s negative, by definition, because there’s almost no profit. There’s no revenue, there’s no growth. It takes a while for the compounding interest, the exponential effect to take flight.
So we have to develop a new discipline, a rigorous discipline that’s just as good as traditional accounting, but that can apply during that long flat part of the hockey stick when traditional metrics are all zero.
Morten: We were looking at a large printing company, and they were trying to get digital printing off the ground, as opposed to traditional printing. They were getting some traction in terms of developing the service, but their numbers were flat. Then the division president says, “If you don’t show me the numbers two quarters from now, I’m going to shut it down.” Sure enough, after two quarters, the numbers weren’t there, and he shut it down.
That was the end of the project. We need different kinds of metrics and accountability systems. Is there one or two metrics that you could propose that are milestones of progress instead of financial numbers?
Eric: I would say the key idea is to focus on the leading indicators of the financial impact that you want. So the trailing indicators like profitability, revenue, growth rates—that tells you how you did in the past. The leading indicators are things like conversion rates per customer, revenue per customer. In your printing example, we may not have a lot of revenue, but can we show that we’re learning how to get revenue per customer up period after period, even though the number of customers are small? That’s the flavor of innovation accounting.
Morten: So there are a number of these structural changes that need to happen. If you’re a manager running a fairly small business, where do you start?
Eric: You’ve got to treat what you’re doing today as an experiment. So if some of these ideas sound intriguing, don’t just blow up the whole organization and change everything. Try to think intelligently about “What’s an experiment I could run today to see if these ideas really make any sense,” and start that learning loop today. Try something, make an adjustment, try again. Over time, gradually allow the transformation to grow until it encompasses the whole organization.