How to Avoid Failing With Lean — It’s a System Mindset Problem
Why do large companies struggle with Lean Startup? The answer to this question can often depend on where you are considering the problem from.
Many experts have tried to address the question already. Some, very well, like Steve Blank in his recent interview with the MIT Sloan Review, others, not as much. But one thing is certain… I will not bore you with another shot at trying to do so.
Instead, this article is devoted to providing a more pragmatic approach to what large corporations must do to avoid failing with Lean Startup. While many have pointed at different reasons why the methodology has failed in large companies, such as GE, there’s still a consensus that it is valuable and effective, when correctly applied.
The Lean Startup works for successful startups, no one can deny that. The question now is what larger organizations can learn from this and implement to jumpstart their own innovation processes. That’s what I’m going to explain in this article.
Why Lean works so well in successful startups
The Lean Startup mindset was created specifically with startups in mind. According to Steve Blank, the philosophy stemmed from a realization that startups are not smaller versions of large companies. While this sounds simple enough now, hardly anyone was thinking that way.
Since startups are inherently different, they require different tools to achieve the unique outcomes only they are suited to. So, Lean provides startups with what they need to achieve success with limited resources and smaller teams.
But what makes the methodology so powerfully effective in successful startups boils down to two important factors. The first is that these startups, as described by prolific technology venture capitalist Gigi Levy-Weiss, are learning machines. The second is that through their iterative learning, they develop a system mindset that lets them get better at every turn and quickly too. Let’s look at each of these.
The learning organization
Levy-Weiss explains that: “there’s an underlying mechanism in iconic startups that often goes unseen: they are learning machines, constantly improving themselves to deliver better results. Beneath each function in a successful startup is a series of feedback loops and a team of learners.” He calls these kinds of organizations learning organizations. You can understand why he calls them learning organizations (as opposed to learning startups) when he subsequently mentions how the Israeli Air Force was the first organization of this kind he had met when joining the military.
So what exactly are learning organizations?
They are organizations that leverage on fast learning to stay ahead of the curve and gain ground faster, compared to their competitors. In a nutshell, these are organizations that improve rapidly between iterations.
The Israeli Air Force has some of the best pilots in the world, yet with a very low amount of training hours. They’ve accomplished that simply because after every performance, a thorough debrief occurs, scrutinizing the wins, as much as the losses. If everything worked well they would ask: “What could have made it better?”
This is a lot harder than just assessing what went wrong in hindsight. For them, there is no such thing as a perfect performance — if you can’t point out any mistakes or say what could have gone better, you are either being blind to your own flaws or afraid to be brutally transparent. And transparency is imperative in a learning organization.
These organizations have built a system of fast-paced, incredibly honest learning through feedback loops. So, what makes these organizations stand out from the rest is, because of these feedback loops, they always start the next iteration better than the last.
The gains of thinking and learning in this manner justifies the effort it takes to implement. As Levy-Weiss states, “assume that in every iteration a learning organization gets 10% better than a non-learning one. The compounding effect post 10 iterations makes it 2.5x faster than the non-learning competitor.”
In summary, here’s what learning organizations do:
- They always look back: Every single performance is debriefed and the outcomes form the basis of the next iteration.
- They are incredibly honest: There are no perfect performances, neither is there space for prevarication. Transparency is critical to learning, so is having a safe, non-judgemental, non-consequential, environment that encourages this.
- They focus on wins as much as losses: Failing to learn from a loss is just as bad as failing to learn from a win (perhaps even more so). When you look back on a win, you can at least find out what you did well and amplify it in the next iteration.
Now we understand how these organizations learn, the next important factor to their success is how they make learning a system.
A system mindset
Gary Kasparov, the fifteen-year undefeated world chess champion, operates under the “learning principle” and describes it as key to his success in his book How Life Imitates Chess. But he doesn’t leave it at that. He goes one step further.
When he analyzes why a move is bad (for example, why ‘pawn-takes-bishop’ loses the game), that is a level 1 strategy or an outcome mindset. After a bad move costs him a game, however, Kasparov analyzes not just why the move was bad, but how he decided on that move in the first place, what was the decision process behind the move.
This is a level 2 strategy, or a system mindset, as described by Safi Bahcall in his bestselling book, Loonshots (which I highly recommend to every innovator). Pixar Studios also adopts the same principle in Ed Catmull’s famous “brain trust” and “dailies” feedback sessions.
So, these successful startups are not only invested in learning from their wins and losses, they are devoted to tracing how they reached that point and picking apart the foundations of success or failure.
As I am sure you are familiar with Lean, you might have noticed already that it operates under the same dynamics of feedback loops and constant iterations. Every customer iteration validates or invalidates a new hypothesis for the business. “We expect this new feature to reduce dropout rates by 12%” or “A wearable, real-time thermometer is the fastest way to identify potential cases of COVID-19 within our field engineers”.
We go to market, experiment, iterate, collect data and validate, invalidate or keep testing. The only thing that is sure is that, before discussing if that feature’s button should be blue or red, we made sure it actually reduces the dropout rate by 12% (the riskiest assumption). Then we start the next iteration “smarter” than the previous one.
Organizations that operate in this manner on all levels are difficult to beat. By constantly soaking up information about what is going well, what is not and managing projects, processes, and portfolios in this manner, they constantly position themselves to strike where the iron is hottest and fire it up where it is cooling. Add a properly implemented Lean Startup methodology into this mix and you can see why it works so well.
How can large corporations learn from this?
Clearly, these concepts are simple to understand, and even to implement, as many corporations have rightly done. Then why are many still struggling to get this right? The problem is one of systems and scale.
Lean works amazingly well, at a project level. It works well when you’re building one startup. It works well when you’re working on a single project in an incubator or even on extended sprints. But when you’re a global conglomerate, with R&D centers across 30 different countries, building a plethora of products in various shapes and sizes, things get a little trickier.
Often, projects that are developing nicely in incubators suddenly lose traction and sink when re-introduced into the core business funding and management processes. Startups are not small corporations and corporations, definitely are not large startups. When you’re trying to manage Lean projects within traditional corporate innovation practices, they run into direct conflict with the current structure and explode. So you need to escalate the model. You need to apply a system mindset at a portfolio level.
Lean Startup is as near a perfect risk-mitigation strategy as any organization can get to. It lets you sift through several ideas quickly and cheaply to find the ones that stand the greatest chance of delivering big results. But you cannot implement this cutting edge methodology with antiquated planning practices. To achieve success, things must change.
The question is how?
In my next article, I will explain how organizations can take advantage of these concepts and create processes driven by Lean Startup techniques that deliver the big ideas that drive growth.