Some Thoughts on the Limits of the Lean Startup

Through various slides and programs, I have so far recommended Lean Startup–style methods. Along the way, I think the limits of those methods have also become visible.

The Lean Startup is a method of gradually building a hypothesis while creating an MVP and conducting customer interviews. In a sense, it can be called a method of "sharpening a hypothesis through customers, together with customers," that is, a method of co-creation with customers.

For that reason, its outcomes depend heavily on the customers. Which customers you choose at the very beginning of a business, and the quality of the feedback you get from them, determine the direction of that business. Watching startups from the sidelines, I also get the impression that who the early customers are changes quite a lot how far the business goes and in which direction. If you meet good customers, you move forward all at once; if you meet bad ones, you are tossed around by their words and only time passes.

So how is the quality of customer feedback determined? It is determined almost entirely by the kind of work and life those customers usually have. For example, in B2B, you can hear advanced needs from companies that are doing advanced things. On the other hand, from companies that have been slow to adopt digital technology, you should be able to hear problems related to that.

The words of Samuel Smiles in Self-Help, to the effect that "the level of politics is a mirror of the level of the people," seem to apply to the relationship between customers and new businesses as well. Perhaps we can say that "the level of a new business or startup is a mirror of the level of its customers."

Why the Lean Startup Produces Global Companies in Silicon Valley

So why are global startups being born in Silicon Valley using Lean Startup–style methodologies? I think the answer, too, comes down to "the quality of customers."

When you start a company in Silicon Valley, the customers around you are basically global companies or companies that will become global companies in the future. By optimizing for the customers in Silicon Valley, a startup finds that even if those customers are small now, they will eventually go out into the global market, and as the customers grow, the startup's problems can grow with them.

That is why, I feel, it is specifically in Silicon Valley and the United States that Lean Startup–style methods can lead companies into the global market.

In particular, looking at things like the excessive development of SaaS in the United States, I especially feel that "the level of a new business or startup is a mirror of the level of its customers."

Policy Goals and the Lean Startup

Recently, in Japan, I have often heard policymakers and educational institutions talk about "producing global startups" and "creating unicorns." But I have a sense that these goals do not combine well with Lean Startup–style methods.

For example, if a startup wants to develop a global business, it will need to win customers who do business globally, or customers located overseas, as its early customers and draw them into the feedback loop of improvement. But not many Japanese entrepreneurs have such a network of customers from the start.

Especially when the resolution of an idea is low, customer interviews inevitably start with the people closest at hand. As a result, many Japanese entrepreneurs meet again and again with Japanese customers who focus on the Japanese market, run their hypothesis testing among them, and inevitably end up with a business optimized for Japan.

The same is true of B2C. If you try to test hypotheses through Japanese customers, there are more customer segments that respond to low prices than to high-priced added value, so you will end up making products tailored to those customers.

Of course, there are also problems shared around the world, so it may be possible to imagine such problems and build products and businesses around them. But human imagination has limits, and that is precisely why we are supposed to co-create with customers using methods like the Lean Startup, so the probability of such big shots succeeding becomes low.

Of course, in Japan too, the situation is different for startups targeting manufacturing and similar fields. Japanese manufacturers often deal with the global market, and in that case you may be able to encounter global problems within Japan. Also, in some areas such as healthcare and pure-software B2C, if you can capture problems common to the world, you may be able to expand globally even if you start in Japan. Depending on the field, such as space or crypto assets, you need to aim for the global market from the start. With pure software such as Notion or developer tools, you may be able to approach problems that are common worldwide.

However, I feel that if you adopt Lean Startup–style methods in Japan, unless you think and implement them very carefully, the result will basically be tailored to the problems of customers in Japan, and will move away from policy and educational goals such as producing global startups and unicorns.

Ease of Teaching and Its Drawbacks

I think the methods of the Lean Startup and Y Combinator are well articulated, and I think they are still effective. There can also be a strategy of capturing the market in Japan and then going abroad, so it is not that focusing on Japan first is bad. Moreover, Lean Startup–style methods should be effective when solving local problems.

However, if we really want to do what has recently been set out as policy and educational goals, my current thinking is that we need to come up with a methodology different from Silicon Valley approaches such as the Lean Startup.

But I have not found such a methodology, and even if there were one, it would hardly be highly reproducible.

And when it comes to teaching, it is troubling that Lean Startup–style methods, whose methodology is somewhat well established, are easier to teach. This is so even though, as a result, startups that solve problems close at hand become more likely to emerge.

English education and sending people overseas will probably be effective in the long run, but in the short run I feel they will not have much effect. If the customers in front of you are Japanese customers, you will be pulled along by the voices of those customers.

So a different strategy or method is needed. For example, I feel it may be necessary to create an environment where founders can meet prospective customers at global companies from Day 1. Honestly, though, I think that is quite difficult.

Alternatively, rather than a Lean Startup that discovers latent problems, perhaps we need to reconsider whether to produce startups that properly solve already visible global problems with technology. That is what I have been thinking a little.