Love the problem, not your solution

Last month, Vlerick Business School together with Groeiatelier, an independent Ghent-based knowledge centre run by and for entrepreneurs, organised a two-day workshop entitled “Running & Scaling Lean: Life's too short to build something nobody wants”. It was led by best-selling author Ash Maurya and hosted at our campus in Ghent. Professor Miguel Meuleman seized the opportunity to ask him about his work.

Miguel Meuleman and Ash Maurya
Miguel Meuleman (Professor Entrepreneurship, Vlerick Business School) and Ash Maurya

Miguel: You’ve developed the Lean Canvas, an adaptation of Osterwalder’s Business Model Canvas and written two books, “Running Lean” and “Scaling Lean”. Most of us have come across Eric Ries’s “The Lean Startup” or its theories in some form or other. How does your work add to that?

Ash: What I’ve found over the years is that, in addressing the challenge of how to build more successful products, faster, The Lean Startup is interesting, but not enough. It has this mantra “get outside the building with a minimum viable product”. But the problem is that more often than not it turns out this MVP invalidates your idea because you realise you won’t be able to create a big enough business, you’re not addressing the right customer or you uncover all sorts of problems. If you’re lucky, you can pivot, but all too often you just get stuck. Now, there’s no point in spending weeks building a MVP only to realise people don’t want it. So, in my more recent work, I’ve tried to answer what comes before the MVP, and I believe it’s business modelling.

Put differently: The Lean Startup is very effective at testing ideas, but the Achilles heel is that these ideas have to be good ones. Which raises the question: where do we get them from? And that’s where a lot of techniques from design thinking, jobs to be done and the basics of good interviewing come in. So, much like a scientist, you have to start with a model, break it into assumptions, and then The Lean Startup helps in testing them, using the build-measure-learn loop.     

Miguel: In your opinion The Lean Startup doesn’t sufficiently address the business modelling phase? Is that what you’re saying?

Ash: In a way ... The methodology says you should come up with assumptions, but there is not a lot of guidance on how to do that. That’s why my advice is: before building a solution, do some business modelling – which is where the Lean Canvas is helpful. So, do some problem validation, and only then build a MVP. Because not only do MVPs take a long time to build, I’ve found they also take a long time to validate. If you can validate faster doing problem validation interviews, than that’s a better approach.

Miguel: Would you then say that your approach is more data-driven?

Ash: It’s data-driven, but the earlier stages are qualitative rather than quantitative. The Lean Startup prescribes we should build products like we do science experiments. But a truly scientific approach requires a lot of rigour, what with double blind testing and all. Now, you can’t possibly tell an entrepreneur that they have to go and find customers, but that they can’t qualify them or get any information about them. That would be tying their hands behind their backs. Entrepreneurs are driven by producing, as quickly as possible, business model results that are repeatable and scalable. And to do that you can’t always take the scientific approach. So, while my approach is data-driven, I’d say you should be looking for signals in the noise and then double down on those signals.

Miguel: I see ... And it’s not as though entrepreneurs can go out interviewing hundreds of prospects. So, drawing conclusions from the data they do collect is bound to be difficult. How do you overcome the many biases such as sample selection or confirmation bias?

Ash: It’s one of the things I address in my more recent work. The key to avoiding confirmation bias lies in creating good interview scripts. Regarding confirmation bias, questions such as “if I built this, would you try it?” are extremely biased because interviewees would say “sure I’d try it”, and you would think they’ll buy it. Asking the right questions helps reducing the number of interviews. An interview is qualitative. There’s no point in creating stacks of interview notes or spreadsheets with rankings. At the end of the interview the only metric that counts is whether you made a customer or not.

Miguel Meuleman - Vlerick Business SchoolMiguel: But how would you define creating a customer in the very early stages, say the problem discovery stage?

Ash: During problem validation interviews, you should be looking for evidence that, in the past, people have already tried to solve this problem. You’ll want people to tell you about the products they’ve tried but didn’t work, about the problems they ran into, so that you can go and design a hopefully better product. When you get to the solution interviews, that’s where the rubber hits the road, because then you’ll be showing a demo of your product and people either sign up or they don’t.

Miguel: Your second book, “Scaling Lean”, is said to draw heavily on systems thinking. Now, systems thinking is one of today’s buzzwords, so, I’d like to clarify: how has it influenced your work?

Ash: I use this metaphor of “the customer factory”, because we’re in the business of creating more customers. Now, the Lean theory has its roots in manufacturing, and in its simplest form it’s about reducing all sorts of waste. But when I look at start-ups, there is waste everywhere, so the challenge is rather prioritising which waste to reduce. And that has been discussed in this other book Theory of Constraints” by Eliyahu Goldratt, which is where the system’s thinking influence comes in. The basic idea is that if you only improve your bottleneck, you improve your entire business. A business model is similar to a factory in that it’s a series of steps and there’s always a bottleneck to improve. In the beginning it’s the problem definition. If you don’t get this right, than building a MVP is waste, because it may result in unused work. So, first, show evidence that there is a problem, then go to the MVP and from there you figure out the next bottleneck in your business model.

Miguel: Such as? What are other typical bottlenecks? HR, Sales?

Ash: Initially, bottlenecks are probably related to the product, so you’ll have to improve your product to improve the customer experience in order to acquire more customers. But at some point it becomes more of a scaling challenge, which could be related to personnel. You mention Sales. Sales reps are good at executing sales plan, not at drafting them. As founder you’ll have to sell to the first 10 customers yourself, because in doing so, you’ll learn a lot about how to sell your product. Growing from 10 to 100 customers isn’t something you as an executive can do on your own, so that’s when you need to hire a sales team and train them. But only take that step once you realise you’re on that threshold. Don’t even think about bringing in a VP of sales and a sales team until you’re ready to move from 10 to 100 customers.

Miguel: Another bottleneck for start-ups, that are initially successful, is market size. They start out in a niche and once they are successful in that niche they want to conquer the world. Is this something you discuss in your work?

Ash: I find European entrepreneurs tend to think too locally, i.e. because of cultural or linguistic barriers they make certain choices, for example about language, and then sooner or later they’re confronted with the challenge of market size. One way to avoid this bottleneck is to think globally from day one, i.e. their back-of-the envelope calculations for the next five years should be on the assumption that they’ll be going global, and then they can incorporate that idea into their business model right from the start. Alternatively, they could make sure to focus on a use case or a job to be done that’s universal. Amazon’s universally appealing value proposition is to provide the best products and good prices. Or pick a local problem, but one that occurs all over the world, like Uber did.

Miguel: Is it really important to think globally or generically from the outset? Can’t you adapt your model along the way? What about focus?

Ash: In the beginning you have to be focused on your early adopters and their niche market, but as you move in that market you’ll begin to see patterns emerging to which you can adapt. My first book was a side project, but talking to entrepreneurs I soon realised entrepreneurs around the world fear the same things and make the same mistakes. So my framework, which I thought would work only in a small niche market, has been generalised over time. Similarly, in just two years the Lean Startup movement grew from a few to 190 meeting groups worldwide, discussing the Lean Startup principles. When you see something like that, you know you’ve hit the nail on the head, you’ve identified a problem people are hungry to solve. So, yes, you can adapt over time. Make sure you’re looking for patterns that indicate a general need.

Miguel: The Lean Startup, Running Lean, Scaling Lean ... There are lots of books, workshops and discussion groups. Is the method actually applied by practitioners?

Ash: Yes, I’d say there is a widespread use of Lean Startup techniques and of the techniques I’ve described. Unfortunately, to really apply the Lean Startup method, you have to be very rigorous and disciplined. So in the past, teams sometimes would cherry pick only some of the techniques. For example, they’d be very serious about it up until the MVP, but once they’d passed that hurdle it was business as usual and they would go back to their old ways. This said, I find that nowadays entrepreneurs are starting to apply the process more rigorously, as there is this growing awareness that the alternative, to just go out and build something, could turn out to be a costly guess. The same holds for investors. They like these methods and techniques because they help to reduce some of the risk in the early stages.  

Miguel: I’m afraid in Belgium, and Europe for that matter, early stage investors are only beginning to catch up. At least until a couple of years ago, I had the impression this discipline was still lacking, also judging by the metrics they were using. Their dashboards would feature mostly lagging metrics, such as market share or sales. But these only record past performance.

Ash: That’s why I’m more interested in leading indicators, which point to future performance, i.e. the ones that show a start-up can make more money six or nine months from now. Entrepreneurs have to tell a compelling story as well as demonstrate progress, or growth. I must say, in the US, investors have a clear growth mind-set. In the past, start-ups could get away with presenting a single good month that was due to exceptional circumstances. They’d rush to raise money, but no sooner had the investors got on board than the numbers went down.

Many investors, even in the US, have learned their lesson the hard way. They’re now looking for proof that a start-up has insights in how to unlock growth hacks. You know the famous hockey stick curve? It’s usually drawn as a smooth graph, but actually it’s a step function, a series of steps. I like to use the rocket trip analogy. In the initial validation phase, a start-up has a handful of customers. Growing is like taking a rocket trip and aiming for ever-higher orbits. Firing rockets will get the rocket, and the start-up, from one stable orbit or growth step to the next. But each of these firing rockets eventually burns out and needs to be replaced with a new one. You could think of these firing rockets as growth hacks.

Miguel: So these growth hacks are in a way removing the bottlenecks in your business model?

Ash: Exactly. As a start-up you’re going to double down on where the business model is holding you back. And that’s what the conversation with investors should be about: how do we show growth in the model?

Miguel: One of 10 key principles you describe in your Bootstart Manifesto is “Love the problem, not your solution”.

Ash (nodding): Yes. As entrepreneurs and customers we still keep making the same mistake, i.e. we gravitate towards solutions too quickly.

Miguel: Incidentally, I’ve had this discussion recently with some of my students. They argued, not entirely without reason, that most of them are not trained to find and understand problems and that our educational system essentially teaches them to come up with solutions. Montessori schools seem to be among the few exceptions nurturing this inquisitive mind-set you also advocate.

Ash (smiling): There was this article in the Wall Street Journal on the Montessori Mafia, which pointed out that people like Jeff Bezos, Larry Page and Sergey Brin had all gone through that kind of education. Mind you, it’s not for everyone. It values independence, lets pupils choose their work and tries to play to individual strengths. Traditional schools tend to try and level everyone ... But things are changing. In the USA a lot of high schools have started to organise start-up workshops where they teach students for example how to work with the Lean Canvas. What I like about these initiatives is that they’re not so much about creating the next Facebook, but about instilling this mind-set that there isn’t a single right answer, that there are lots of unknowns and uncertainties.

Miguel: Ah, such a shame you don’t speak Dutch, then. Lieven Scheire, a popular stage and TV personality has just launched a call for problems. He’s preparing a new television show for the Flemish public TV station, in which all sorts of problems will be submitted to a panel of geeks, inventors and scientists. You’d probably love it.

Related news

  1. The success paradox

    Date: 05/11/2018
    Category: Opinions
    The knowledge, experience, image and sincere concern of a founder are irreplaceable. However, a study of internet companies by Harvard has demonstrated, ironically enough, that a founding CEO’s success in achieving important company milestones increases that person’s chances of being replaced as CEO. Opinion article by Veroniek Collewaert, Professor of Entrepreneurship.
  2. AI: Science fiction or not?

    Date: 14/08/2018
    Category: Opinions
    These days, artificial intelligence is everywhere and both large, established companies and young startups are busy experimenting with all kinds of AI applications, across all sectors. Science fiction? Actually, a lot of it comes down to advanced data analysis. Some people also call it “statistics on steroids”, as AI largely tries to recognise patterns in data. Legislation will play an increasingly important role here, a point which should not be underestimated if we do not wish to fall behind other continents that have already made greater progress. Professor Veroniek Collewaert is inviting politicians to already respond to the impact of all these evolutions on our society and to find the best ways of preparing for this.
All articles