In his book Zero To One, Peter Thiel has some harsh things to say about lean startups. He writes that lean is “code for unplanned” and equates the Lean methodology to “making small changes to things that already exist.”

He argues that, “would-be entrepreneurs are told that nothing can be known in advance.” He characterizes customer development as nothing more than “listening to what customers say they want,” and says that Lean tells founders to “Make nothing more than a minimum viable product and iterate [their] way to success.”

It’s hard to read Thiel’s critiques and avoid the conclusion that he is reacting to the buzzwords that surround Lean instead of actually thinking through the concepts behind it. This is a shame, because Thiel’s success gives his proclamations tremendous influence. When he’s wrong about something important yet speaks confidently about his position, a lot of people will follow him.

It’s a shame because if he actually took the time to think through the context, Thiel could add a lot to the conversation. And as someone who claims to be serious about the practice of philosophy, he should know better.

The thing is: The Lean methodology is brilliant, and if more entrepreneurs understood it and applied it rigorously, fewer startups would fail. It might even save a few endangered unicorns from the axe.

The Wrong Assumptions

Lean assumes that on day one, founders start with a series of untested hypotheses about their potential customers, the ideal pricing for their product, their best distribution channels, the costs of operation, and so on.

Rather than simply assuming you are right about all of these things and can dive headfirst into building your product, Lean suggests that you test your assumptions first.

In case you’re not familiar with the methodology, it (more or less) goes like this:

Step 1: Map your assumptions on a business model canvas.

Before you write a line of code, build a piece of hardware, or develop a new pharmaceutical compound, you get together with your co-founders and map out all of your core assumptions about your product and the market for it:

• Who will be your buyers and end-users?
• What huge problems and/or desires do they have that will make them want and need your product?
• Who are all of your potential partners? What can you offer them?
• How will you distribute your product to the market once you’ve built it?
• What are the biggest obstacles in your way?
• Who are your most important competitors?
• What regulatory agencies might take an interest in you if you get big? How will you handle them when they show up?

And so on.

Building a business model canvas is the definition of planning. It is not, as Thiel would have it, “code for unplanned.” Encouraging founders to define and test their assumptions is also the opposite of telling them that “nothing can be known in advance.”

Strike one.

Step 2: Get out and do customer development

Big, exciting ideas and visions about your product and your market are all well-and-good, but sometimes (often times), they are flat-out wrong. Indeed, they are only hypotheses.

According to Lean, it is both practical and cost-effective to test these hypotheses by getting out from behind your laptop, going out into the real world, and talking to flesh-and-blood human beings.

This process is known as “customer development.”

Contrary to our famous contrarian’s beliefs, customer development is not a “focus group.” It doesn’t even mean “asking potential customers what they want.” While asking potential customers what they want is almost definitely something most founders would benefit from doing more often, it not the whole shebang. You learn a lot by defining and talking to your potential customers.

Indeed, customer development is one of the lowest-cost ways to test your startup’s core assumptions and validate or invalidate them before getting all hot-and-heavy building your product.

The purpose of customer development is to inform your vision, so you’re not building blind. If nothing else, this makes you much more likely to build something people want. But instead of dissecting this concept, Thiel writes the whole process off as a “focus group.”

Another swing and a miss.

Step 3: The pivot and the minimum viable product

One of the big buzzwords around Lean that gets thrown around with reckless abandon is “pivot.”

But most of the people talking and writing about pivots don’t understand what they are. A pivot is not when you fire most of your team and start a new product from scratch.

For example, Stewart Butterfield did not “pivot” from his failed flash game, Glitch, to Slack. He simply failed and moved on. The founders of Turntable.fm did not pivot when they abandoned Stickybits and started Turntable. They simply killed Stickybits and built a new startup on top of its remains.

In startups, you pivot when you stand firmly on your foundation and adjust and readjust your concept, vision, and product roadmap until you find yourself with something that the market actually wants.

This brings us to the minimum viable product, also known as the “MVP.” Peter Thiel frames an MVP as nothing but a half-baked product that you launch into the market and then iterate on until you succeed. But again, he gets it wrong.

Much like customer development, an MVP is simply a way to rapidly test your hypotheses, deepen your understanding and clarify your vision. As Steve Blank says, “an MVP is whatever you need to show customers to maximize your learning at that point in time.”

Unlike a slide deck that explains your idea in the abstract, an MVP makes it concrete, giving potential users and customers something they can wrap their minds around. Their reactions—positive or negative—help you understand what will work and what will fall flat, making it more likely your end result is a product people will use.

This is very far from (to use Thiel’s framing) making “nothing more than a minimum viable product” and then “iterating your way to success.”

Strike three.

Why Lean Is Important

Contrary to these misguided characterizations, the Lean methodology is the closest thing to a scientific approach to startup creation that we have.

That is a bold claim, so let me back it up: The scientific method boils down to using your intuition and understanding of the world to develop hypotheses and then rigorously testing these hypotheses against reality to determine if they are true or false.

Of course, in order for you to make a claim on the truth, the tests you use and the results they generate must be repeatable by others — and not just those you’ve paid off. In the world of startup creation, the Lean methodology doesn’t quite achieve the rigor of repeatable tests, but it is certainly one of the most effective, cost-efficient ways to develop and test hypotheses about your product and your market.

If more founders understood lean and practiced its teachings before diving in and building and launching products, the startup landscape wouldn’t be as littered with endangered unicorns, zombies and acqui-hires as it is today.

As Steve Blank said when I interviewed him on the subject:

“The goal of a Lean Startup is to minimize resources, time, and dollars, and maximize impact. I don’t know how you could disagree with that, but the best I can fathom is that Thiel has a visceral dislike of the word Lean. Given Thiel’s intellectual horsepower, it would be a lot more interesting if he contributed his skills to really understanding and extending the strategy.”