People forget that the defining characteristic of the hockey-stick shape is the long, flat blade, not the vertical shaft. I’ve been in startups where we launched our product, and six months afterward we thought we were on track. “Wow, we have almost no customers and almost no revenue; we’re on plan. We’re going to have a hockey stick in three, two, one....” Then I had the really painful experience of being in the flat part indefinitely. There was no meteoric rise up the handle.
What happened? We were on time. We were on budget, on schedule. We did everything that we said we were going to do. It took me a long time to understand the real fundamental issue: It is impossible in a startup situation to make an accurate forecast.
Then how do you measure your progress? How do you hold yourself and your colleagues accountable during the early stages of building a new business? This ties to my greatest fear as an entrepreneur: going to sleep at night unsure of what I accomplished. I know I spent some money. I kept a lot of people busy. We built some new features and hit some development milestones. But if I’m building something that nobody wants, why would I be proud that I did it on time and on budget?
The antidote is validated learning. You prove one step at a time that you are figuring out how to build a sustainable business. That’s where the concept of minimum viable product comes in. Instead of building the whole product over many years and shipping it all at once, you try to find the smallest experimental version of the product you can launch, to begin the process of getting feedback.
S+B: What are you measuring in that feedback?
RIES: Learning is the unit of progress in entrepreneurship. It’s more important than making money, getting customers, building features, or engineering technical quality. Of course, those things are important, but only insofar as they contribute to learning what creates value and what creates waste.
In the lean startup, we assume that we know nothing; we don’t even know who our customer is. We have nothing but a hypothesis. We use that hypothesis to pull experiments through our “factory” as quickly as possible and get them into the hands of potential customers, with the goal of maximizing learning and eliminating waste. At the outset, waste is anything that doesn’t contribute to the learning.
S+B: How might an actual product go through this process?
RIES: Our first goal is to build a minimum viable product: a working prototype we can get into the hands of prospective customers. We do this by constraining the production run—say, to 100 or even just 10. Then we move right past traditional market research; we just go ahead and sell them. We skip traditional distribution channels, which would obviously be wasteful for such a small number. We go to one store, or maybe one customer, and try to get them to use the product. We sell it ourselves. There’s no sales team. We are personally in the store persuading customers.
S+B: It’s like something out of a 1935 county fair.
RIES: Right, it’s extremely low-tech. The goal, remember, is to start the learning process. I don’t care if you’re a new company or P&G—chances are, you won’t succeed the first time you try. For example, if no one’s ever seen a Swiffer mop, how do you know what aisle it should be in? All sorts of questions come up with a truly new product. People walk up and say, “What is this thing, and why are you trying to sell it?” Most people won’t be interested.