Steve Blank Hands A New Owner’s Manual to Startup Founders
If you look closely at the cover of Steve Blank’s new book, The Startup Owner’s Manual, you’ll see that it shows an exploded view of the transmission from an automobile engine. Leonardo da Vinci invented the exploded view more than 500 years ago, and it has stayed with us as a visualization tool, because it’s great for showing relationships between the parts of an object. But it’s also utterly unrealistic (as many of Leonardo’s fancies were). In the real world, once you’ve taken apart something as complex as an engine, you’ve usually got a huge mess on your hands, and no obvious way to put it back together.
Startups are sort of like that. You can boil down your dream for wowing customers, crushing competitors, and taking over the world into a few orderly spreadsheets and PowerPoint decks—and you might even be able to get a few VCs to buy your pitch. But making your plan happen is another thing entirely. As Blank likes to say, no business plan survives its first contact with the real world.
Read enough of Blank’s books and attend enough of his talks, as I have, and you begin to see the inescapable truth of his arguments. A startup is not a business, he argues. It’s a machine built (and often rebuilt on the fly) to find a viable business as quickly as possible, preferably before the fuel runs out. Once it’s found that business, it can graduate to being a real company and spending real capital. In the meantime, it’s all about the search. And that’s the process Blank has been thinking, writing, lecturing, and teaching about for years.
Blank’s first book on entrepreneurship, written six years after his marketing-tech company E.piphany raised $66 million in a 1999 IPO, was called Four Steps to the Epiphany, and you’ll see sticky, dog-eared copies of it on the shelves of many serial startup founders. It codified “customer development”—the idea that a startup’s job is to turn its hypotheses into facts by listening to real customers about whether they’d use a proposed product and what features they really need. But that book, not to put too fine a point on it, was a turgid, rambling, difficult read—and a lot has happened in the world of entrepreneurship training since then. For one thing, there was the emergence of Eric Ries’s “lean startup” philosophy, which argues that companies should apply the rapid-iteration method from agile software development to their entire business. Then there’s the business model canvas, Alexander Osterwalder’s popular method for visualizing a startup’s value proposition, customers, and revenue streams.
Blank has been incorporating all of these ideas into the entrepreneurship courses he’s been teaching for Berkeley, Stanford, and the National Science Foundation. But there was no updated textbook that laid out Blank’s full entrepreneurship curriculum. That is why he and co-author Bob Dorf, a Connecticut-based serial entrepreneur, have now published The Startup Owner’s Manual: A Step-by-Step Guide for Building a Great Company.
The book is long—573 pages—but it’s far more organized, methodical, and textbook-like than Four Steps, and is therefore likely to be snapped up by B-school teachers and students, as well as entrepreneurs. The 12 chapters are broken into two big sections, one on “customer discovery”—the process of testing the founders’ vision against the problems and needs of real-world customers—and one on “customer validation,” where you scale up the product to see if it’s got the potential to be a big business. The book has material for every type of entrepreneur, but it helpfully highlights the sections that may only pertain to the crazy world of Internet startups. (Sections about selling physical goods are in serif type, while sections with advice for Web and mobile startups are in sans serif.)
I confess I didn’t read The Startup Owner’s Manual from cover to cover before writing this review and Q&A. But that’s okay—Blank gave me permission not to. “This definitely isn’t a book,” he says. Rather, it’s an actual manual—something you should dip into as needed. “Reading it any other way, you will get frustrated and exhausted,” Blank says.
I talked with Blank about the book earlier this week. Here’s a version of our conversation, edited for length.
Wade Roush: The Startup Owner’s Manual is extremely detailed—I’d almost say exhaustive. How long have you been working on it?
Steve Blank: I decided to do this, theoretically, the first time I got an e-mail about a typo in the first book. I have been mentally storing things away since then. But how I work may be different. I pay attention to a lot of stuff, and it’s not until a pattern fits that I finally make a move. The real problem in this case wasn’t the typos, it was trying to understand what was going on with Web and mobile startups. The more distance I got from Four Steps, the more it seemed like [building a Web or mobile company] was a truly different process from traditional customer development. I was trying to figure out how that was going on. But it became clear gradually that it wasn’t a different process—it was a different speed, with different channels. What took so long was struggling with the “unified field theory” of startups.
WR: What is that theory?
SB: The more I thought about what we and Eric Ries and Alexander Osterwalder were building, the more it seemed we are truly describing a new way to think about why startups are different. Large companies execute known business models. Startups search for business models. The mistake we have been making, both in entrepreneurship and in entrepreneurship education, is to focus on business plans and organizational charts. Those are execution tools. But we shouldn’t be executing on Day One—we should be searching. Having that insight helped me to understand that, just like an MBA program has a “management stack” of books on executing, we needed our own stack for search.
There is no right process for this search, but once we have the language we can now have the discussion. The mistake we were making was thinking that startups were just small versions of large companies. That was the fatal error, and we were getting it wrong time after time. People who run big companies aren’t innovators, they are financial people, with different risk profiles. To expect accountants and venture capitalists to come up with innovative ideas is an oxymoron—with an emphasis on the second part of that word.
WR: The book is way too long and detailed to be consumed in a linear way. How do you think people will actually use it?
SB: Unlike Four Steps, where I fooled myself into thinking that it was a book, this definitely isn’t a book. It’s a manual. Reading it any other way, you will get frustrated and exhausted. It’s like pointing a fire hose at your head and having someone say, “Here, let me show you everything you are going to spend the next four years doing.”
People can tell you things forever, but the most important things are experiential. You can talk about customer development and getting out of the building, but until you have actually screwed something up, you won’t understand. It would be like reading an auto repair manual without having a car. It’s one thing to read a car repair manual, but let me tell you, when you try to rebuild a carburetor, you wind up with pieces all over the garage. That’s why there is a picture of a transmission on the cover.
WR: You break down the process of searching for a business model into steps like customer discovery and customer validation, and you break down those steps into many substeps, each with their own flowcharts and business model diagrams. It seems to me that an entrepreneur really tried to go through each step carefully, it could take a very long time. How long do you think it should take to launch a startup?
SB: That really depends on market timing. If you are in an existing market and you happen to be a domain expert, then it can happen quickly, in months. If you are resegmenting a market, or you are testing a first-time hypothesis, or you are just wrong, then man, it is going to be a tough first year or two. Which it normally is—remember that most startups fail. No one says that enough. If we were honest about this book the title should really be “Read this and you will fail less.” It doesn’t guarantee that you will succeed more, but it means that you will last longer and do fewer stupid things.
And you will also be acutely aware than when a customer grabs you buy the collar and says “You can’t leave until I have this product,” or when 10 or 100 times more customers show up than you planed, you can put the book on the shelf and go deal with your success, rather than implementing the next step. If that happens, go find the tallest point of your building, tear the book in half, and throw it off the goddamn roof. But the book is for the situations where that is not happening. We forget that 99 percent of entrepreneurship is learning from failure to failure.
WR: One of your most fundamental messages is that startup founders need to “get out of the building” and talk with prospective customers. But most high-tech entrepreneurs are probably the kinds of people who would prefer just to sit at their computer or their workbench, build a better widget, and have the world show up at their door. How do you force yourself to get out of the building?
SB: I’ve got to tell you, personally this was the hardest thing for me, ever. I was incredibly introverted—my default is sitting around the ranch and reading. It explains why startups are best done by teams. But even the most technical founder, unless they want to be an employee rather than a real founder, needs to get out of the building and talk to people, just so they can appreciate how hard it is. Otherwise, if you hire this out, you will always be beholden to someone else who is capable of bullshitting you.
WR: If you don’t have any customers, how do you find people who are willing to take the time to listen to you about your idea and give you honest feedback?
SB: I’m just going through this now with the scientists participating in the NSF iCorps program. Their average age is 45 and they’ve been sitting in the lab for the last 20 years. So the first thing I say is, “Do you have any peers?” “Yeah.” “Got any former students in industry?” “Sure.” So the first thing is to make a list of those people. The mistake is that people think the first meeting has to be with the highest-level person you can think of, but right there is where they go off track. You want to be talking to the lowest-level people, because your odds of getting your presentation right on Day One are zero. You are going to be wrong. And you want to get it right when you get a shot with the CIO or the VP of engineering, so you have to practice.
If you want to get somebody to talk, my favorite thing to do is to call after 6:00 pm, when their secretary is gone, and say “I got your name from a person who said you were the smartest in the industry. Can you give me 15 minutes? I have nothing to sell you, I am just trying to understand if this is an important problem or not.” If I said that, wouldn’t you at least listen to the next 10 sentences?
The hard part is to ask the right questions, and then to shut up and listen like you really do care about what problems they have. What you are trying to do is understand the fit between the value proposition and the customer segment. If you remember the business model canvas, this is just one box—there are seven other boxes. But once you start these dialogues going, customers will teach you all this stuff. How do you prefer to buy? What do you read? How do you learn about new products? Customer development is a process—it’s the z axis of the business model canvas. Once a week you can use what you learn, create a new canvas, and tear off old one. Then later you can play it back like a filmstrip, and holy cow, now we can see the journey!
WR: It’s funny—when I interview entrepreneurs, I always ask how their ideas evolved and what kinds of pivots they went through. But a surprising number of them claim that their idea was fully formed from the beginning.
SB: Entrepreneurs always tend to reinvent the past. They say, “Oh, I had this idea from Day One.” I think the real story is much more interesting. It’s not about how stupid we are, it’s about how well we learn. The ones who get rich are the ones who figure out that startups go from failure to failure, and it’s about how quickly and cheaply you learn from those.
In the past, the whole system we had in place required firing executives to bring about change, instead of saying “Hey, this isn’t the right customer segment.” I used to get called in by some VC friends who would say, “Steve, we have a company with a marketing problem.” I would say, “Why is it a marketing problem?” They would say “Sales isn’t making their numbers, so it must be about positioning.” It wasn’t about positioning. They just didn’t have the tools or the language to know that they weren’t going after the right customer segment. But once they’ve hired 12 people in sales and jacked up their burn rate, it’s too late.
We now have a radically different way to think about this stuff. It means you never look at startups the same way again.
WR: So you’ve got a unified field theory of startups. But as Karl Popper taught, for any theory to be truly scientific it has to be falsifiable—there have to be built-in ways to prove it’s wrong. In what ways is customer development falsifiable?
SB: Simple. If you do this and you actually have a higher failure rate, or you do this and you have no difference in liquidity, it would be false. We now have five or 10 different ways we can measure this. What I loved about teaching the first NSF cohort was that it was pretty well instrumented. They did a ton of surveys with the scientists about their knowledge of customer value propositions and so forth before and after the course. And at least for this small survey sample, we got some really encouraging data. I think what we are seeing is that you can either try to pursue your business through the traditional plan of over-raising venture money and then chasing dollars, or you can chase customers—which is a lot more fun and productive.