I recently wrote about the importance of failure for learning and that as we ramp up to figuring out how to save the world, failure will be a critical part of our path and needs to be embraced. I stated that our culture is too failure-averse for what we need to do and that we need to foster a culture that accepts failure as a natural part of the learning cycle.
I still think that’s true but it’s dawning on me that it’s only the first step towards an appropriate way of thinking about failure.
Our failures must be intentional. Failure is only good when we actually learn from the experience and are able to apply the lessons to the future. I wrote
Think > Do > Reflect > (repeat)
and don’t get stuck in any part of that cycle.
I just finished reading The Lean Startup by Eric Ries, which is my first real foray into the world of lean thinking. Pretty much everything I read in the book struck me as extremely relevant to people and organizations doing worldchanging work.
Failure is a critical component of the Lean Startup model, but embedded in the process of failure is a rigorous process for extracting information from the failures, i.e. learning. A core concept to the Lean Startup model is called the Build-Measure-Learn feedback loop.
Build (a product) > Measure (data) > Learn (ideas) > Next iteration
The key to success lies not in any particular component of this loop, but in minimizing the total time it takes to get through the loop. One of the first steps of a startup is to enter the Build phase as quickly as possible with a minimum viable product, or MVP.
The idea is that one-shot success is virtually impossible (reality is hard to model!) and iteration is expected. The more effort you put into building and polishing a product before testing it against reality, the more effort you have put in that will likely get scrapped. You want to put in just enough features to be able to see how they respond to the variables of reality, so that you minimize the amount of work you do that is irrelevant and maximize the amount of work that gets directly tested and has conclusions drawn from.
Don’t spend time building features you think are valuable into a package and then release the whole thing in one massive launch; test each feature against reality rigorously and be brutally honest about what works and what doesn’t. It’s okay to fake features and services to get feedback about whether or not they’ll actually work before building them.
Vanity Metrics
A common problem with startups, or any productive venture, is a misaligned conception of productivity. A classic mistake is to spend all day hammering through tasks, getting things done and pushing out features. At the end of the day you can feel highly productive, but what did you actually accomplish?
Who cares if you built x widgets if they aren’t actually producing value? You can make metrics for all sorts of things that make it look like you are moving forward (the book calls them “vanity metrics”), but that doesn’t mean you are necessarily moving towards your goals and fulfilling your vision. In fact, you may be getting even further away.
An example of a vanity metric that Ries gives is website hits. What does it actually mean to say that your website got 20,000 hits? Did 20,000 people actually read your website? Or was it one guy with a really hyperactive browser?
An example of a vanity metric in the green built environment, to my mind, is number of LEED projects. It’s popular for design firms to boast that they have some number of LEED Gold projects, some number of LEED Platinum projects. That might be fine to say for marketing purposes and to communicate the firm’s dedication to pursuing green projects, but it’s not a good enough metric to define how “green” your designs actually are.
Sure, it’s great that fourteen of your buildings are LEED Certified – but how much carbon are they saving? How many thousands of gallons of water are they consuming? How much cancer was built into the building? Calling a building LEED Certified hints at these numbers but isn’t actually a decent metric to benchmark your success at building an ecotechnic environment. All of your LEED Platinum buildings might, in fact, use an egregious amount of energy – but if all you’re measuring is the fact that they’re awarded LEED points then you won’t even know that you are actually failing in reality.
In the same way, it’s not good enough to have an idea, build it and push it out, see it fail, and say “ah, well, that failed, it doesn’t work; next idea?”. Why didn’t it work? What caused it to fail? What assumption or set of assumptions didn’t hold up? This kind of knowledge can only come with rigorous measurement and analysis.
In fact, obvious failure without data can be harmful. Often an idea is mostly good but fails due to one missed detail. If the effort isn’t measured and analyzed, people will just see the failure and think that the entire idea doesn’t work and might abandon an entire approach that actually has potential. We can’t afford to throw out good ideas and we can’t afford to have wrong ideas about what works and what doesn’t.
Leap-of-Faith Assumptions
Reis makes the point that at the heart of any new startup lie a number of leap-of-faith assumptions. There have to be; startups are by definition trying new untested things. Entrepreneurs base as much as possible of their business plans on knowledge and observed facts about reality, but down at the core of every startup are some statements that stand alone and are not strictly justified. They boil down to “We think people will pay for X because of Y.”
The task of the startup is to test these assumptions against reality. If the result of the metrics, the validated learning, indicates that the assumptions are false, then it is time to pivot.
“At it’s heart, a startup is a catalyst that transforms ideas into products. As customers interact with those products, they generate feedback and data. The feedback is both qualitative (such as what they like and don’t like) and quantitative (such as how many people use it and find it valuable). As we saw in part one, the products a startup builds are really experiments; the learning about how to build a sustainable business is the outcome of those experiments. For startups, that information is much more important than dollars, awards, or mentions in the press, because it can influence and reshape the next set of ideas.”
I think it’d be really illuminating to take a long hard look at what some of the leap-of-faith assumptions built into a lot of deep green organizations are. I’ll save that for another post.
Succeed at Failure
The entire time I was reading the book I was seeing the parallels between the work of traditional startups and those people and organizations trying to change the world in one form or another (deep green design firms, activist organizations, environmental advocacy organizations, etc).
Startups are trying new things that no one has ever done before. They’re kicked off with a vision of a possible future that has never been tested against reality. They push against the inertia of status quo and try to get massive numbers of people to change their behavior in some way. Startups introduce disruptive technologies and services that change how the world works.
In the same way, no one has built an ecotechnic future before. We don’t really know how it’ll work, what ideas will flesh out and which will turn out to be terrible. I thinkt the tools of lean startups have direct relevance and importance for worldchanging efforts.
What we’re fighting as worldchangers is the massive inertia of the world system, a leviathon, on a one-track trajectory for planetary destruction. The worst possible thing we can do is build our own giant green leviathon, built entirely out of untested assumptions about the world and shoved in one direction based on unvalidated knowledge (myths?).
In summary, my little foray into lean thinking has impressed upon me that we can’t just go out and start failing left and right and expect it to pay off. We must fail well, fail intentionally, fail in a very conscious and reflective sort of way. We need to rigorously measure our failures in a transparent way and develop a specific practice of failure.