Seyi Fabode is fed up with Lean Startup.
While he has respect for the Lean Methodology as practiced by Toyota, an approach he learned while getting his Masters in Manufacturing Systems Engineering, he's now evangelizing about the need for companies to stop striving to apply the Lean Startup approach especially in situations where the company is trying to solve 'wicked problems.'
What's a wicked problem? One that can't be solved by 180 degree-pivots or A/B testing or Lean success metrics. Environmental degradation, terrorism, and poverty - these are classic examples of wicked problems.
According to Seyi, to solve wicked problems you have to go big and engage in systems thinking. You have to abandon what he calls the 'short-termism' of Lean and focus on longer term tests and higher points of leverage.
We sat down with him in preparation for his session at #Innov8rs Atlanta to ask why Lean fails in larger business ecosystems...and exactly what we should do instead.
Your perspective was born out of experience and frustration. What's going on that you think is just not working?
I've worked in the utility industry for about 15, 16 years, doing digital transformations or building products to solve problems that are peculiar to the utility industry. And like many of the bigger ecosystems - energy, healthcare, food and agriculture - the results of actions you take today take a while to play out. There's a lag between action and response. Lean Startup methodology, Lean Business Canvas, all these tools that Silicon Valley and the like have thrown upon startups to use, don't work in these industries.
And for that reason, while there are huge problems to be solved in those industries, they never get the expertise or the money to solve those problems at scale.
I think it's time for people to start paying attention to solving those big, wicked problems with healthcare, with energy, with food and agriculture. But to do it right, we have to use new tools. That's what I'm proposing here.
So if your company is serving millions of patients in healthcare, for example, you're too big for Lean methodologies? MVPing and things like that won't work?
They can, but the results of one test with an MVP won't tell you anything useful.
You don't make decisions like that for ecosystems that are fundamentally shifting. Part of the problem is people doing it wrong because the tools they've been given mean they will not get insightful results.
So, I'm not suggesting that MVPs are useless, but the scale of your problem should determine the scale of your MVP.
I had a client who wanted help figuring out how much customers cared about renewable energy - would they buy solar panels for their homes and energy storage for their electric vehicles. So, they threw up a landing page for six months. They spent about half a million dollars on this landing page with some products they didn't have and asked people to sign up. They got 2 or 3 signups and they were like: 'this won't work'.
A worthwhile MVP for that sort of test of interest, in my opinion, is to equip all the homes on three blocks with solar panels, energy storage for their homes, Nest thermostats or whatever thermostat you're going to use - make the homes fully connected. So about 30 homes, and you probably won't spend half a million dollars. You will learn so much over the course of six months than you will with a landing page. That's how you think bigger about figuring out what works or doesn't work for the ecosystem-based problems we're trying to solve.
If Lean Startup and other methods don't work for these sorts of wicked problems, what will?
We need a long term perspective. Lean Startup methodology and all these tools have unintentionally taught us that we can, and should, get results quickly. I think that is almost as big a problem as using tools that are not fit for your purpose. So expanding your time horizon is the first way to think about this.
More importantly, we need to pay more attention to the results that come from the interactions between all the elements of our tests.
In the example I gave, they were looking for clicks on a website to determine user behavior or user interests. But with a street block, real-life MVP, you can pull the actual data and see that, for instance, one family began only washing their clothes at night because they'd seen the data and it showed them they could reduce their costs.
You can truly see human behavior, which will give you far more information and insight into what business models you can use to serve those customers - what you can offer that gets them to change their behaviour in a way that benefits them, but also makes money for your business. So, expanded time horizons and an openness to results that you didn't think about before…that will inform your strategy.
It sounds obvious. Why isn't it happening?
Innovation and entrepreneurship have become 'Hollywoodised', so to speak. We all think it should look a certain way. And so, even though what we're discussing here is seemingly obvious, the inputs we get from the news, from the technology community, from the people that we look to as innovators, suggest that we shouldn't do it that way. We see X, Y, Z company pivoted from selling one thing to another thing and they've become a billion dollar company.
AirBnb were selling breakfast snacks to people who came for a conference and they pivoted to letting people share their bedrooms. We see that and we're like: 'huh, this is how it's supposed to work'. Well, no. It's not how it's supposed to work for everyone.
The reason those stories and they get amplified is because they are anomalies; but now, we now look at those anomalies and think it's the usual, and try to replicate what they did.
Does that expectation, or that influence, depend on the industry?
I don't have data for this, but anecdotally I think so. The closer your industry is to the products or the services you provide, the closer it is to you and I as consumers, the more likely your industry is to be influenced by what you see a company like Amazon doing. You'll be more likely to think you should be moving super-fast. And I'm not saying you shouldn't, but there's a systems approach to moving quickly.
How would your model help, for example, a large pharmaceutical company? It takes ages to develop products and get them approved. Would it still help them to be faster or more effective?
Yes, and I'll actually share some of the work I've done there during my talk at Innov8rs Atlanta. With this model, one of the things you're improving is the quality of the information you're looking for and the flow of that information across the organization. The better the quality of the information about what's going on with tests or with the organizational hierarchy, the quicker you can make well-informed decisions.
The problem with some healthcare products and services is they are spending so much time on problems that have already been solved, or are in the process of being solved, and it makes their innovation process much slower than it needs to be. For example, insurance companies, hospitals, and pharmacies used to all have different ways of storing customer data. That's becoming standardized now. So you don't need to spend months determining what your data structure should be - you can take advantage of what's already been done, or is being done. You can say: ok, company X, Y, Z used these data structures, so we'll use them too because it allows us to pipe into these companies easily. And then you can get on with really innovating.
You mentioned systems thinking. How do you apply it to get to real innovation insights?
There's no one definition, but basically it's a more holistic way of thinking about business.
Analysis breaks things down - when we analyze, we break things down. Systems Thinking looks at the whole.
We consider the elements of the system, the interactions between those elements, the inputs into the system, and the outputs we get from the system. While there's some analysis required for each of those elements, you never lose sight of the whole.
To apply it, you really just need to understand the underlying perspectives, the views you have to keep in mind. One I use a lot is Leverage Points, also called Points of Intervention - Donella Meadows is a great systems thinker who came up with this.
Essentially it says that for any system, there are about 12 places to intervene within a system to shift it, and they are ordered by effectiveness. The highest on that hierarchy, the most effective, is transcending paradigms. The lowest, or least effective, is numbers.
Whenever we are doing Lean Startup, we're only focused on numbers. But, when you think in the systems approach to things, you start to think about the paradigms, and the culture, and the things that have lead us to the problem we're trying to solve.
Another perspective is around growth. It's a mathematical formula that suggests even the biggest companies - the Googles and the Apples of the world - will reach a point where they cannot grow anymore. When you think this way, you can step back and instead of focusing on Apple making a new iPhone, you question: has Apple hit the limit on the number of iPhones people will actually buy? And if that's the case, maybe they need to shift.
Basically, it's a new way of thinking about and defining what the problem is. It's fun, fascinating stuff.
How does that play out for organizations wanting to create a culture of innovation?
When you can build a learning organization where everyone is keyed into understanding the bigger picture, and seeing their role within that bigger picture, and solving problems as a part of the whole system, the tools they use won't matter much. Well, the tools they use will matter - but they will decide what those tools are.
Say I'm an analyst, and my company is a billion-dollar company. Maybe I don't need Lean. Maybe what I need is an Excel model to help me think about the size of our market, and help me see where and how solving a particular problem in our industry could get us another billion. And then I can reach out to other people and get them to work with me on it. We might still use MVPs and other Lean tools, but when we're dealing with a scale like this our MVPs would need to look very different from what the Lean Startup model currently suggests.
Let me add as a closing note: I share these things sounding like I fully understand them. Thing is, I am constantly learning more about the approach to systems thinking, different perspectives and so on. I'm hoping that during my talk at Innov8rs Atlanta, some people in the audience will disagree with me so that I can learn from them as well. And anyone who reads this and disagrees should say that, so that we can have a conversation and keep learning from each other.
It's how we get better, and how we all move forward.