What I've learned about good prototyping

22 April 2021
22 Apr 2021
New York City
6 mins

The fastest way to move from just an idea to a viable venture is to build a first prototype, or the “minimum viable product”. Over the years, I’ve changed and improved what I think an MVP should be. I think its name leads many first-time founders to build the wrong thing at the start of a company, when resources are most scarce, and right decisions have big payoffs.

I started out thinking of the MVP as exactly what the name suggests – just a minimal version of your product. If you’re building a paid SaaS product, perhaps a minimal version is missing some key features, but more or less works as designed. If you’re building a direct-to-consumer CPG goods company, the MVP might be a basic packaging of the product, and a way to get it delivered to customers.

The problem with this minimal product model of an MVP is that it’s just not what most companies need at the earliest stages. When you don’t have a product or customers, what you need isn’t a worse version of your product – it’s cheap, efficient, fast learning about all the unknowns of the company’s business. Building a shippable product is rarely the easiest nor the fastest way to get the most pressing questions about a business answered.

A better model of an MVP is the minimum viable business definition of an MVP. Product engineering is becoming more and more of a commodity. Building a functioning product, especially in software, and delivering it to users is becoming easier by the day. Increasingly the hardest parts of a business are distribution and differentiation. If you build a great MVP and build something people want, but discover too late you have few ways to get the right people to know about what you made, you may as well have wasted your effort building a prototype. The minimum viable business model of an MVP admits that an MVP for a business includes an MVP for distribution – what’s the simplest, cheapest way for you to reach your customers? How many can you reach? How viable are those ways of distribution? If your business involves manufacturing, a minimum viable business should try to validate your ability to manufacture what you need. If you can hand-build a dozen great products but the same design is impossible to manufacture to the right tolerances or volume, you simply don’t have a business.

At a high level, a company is a bundle of processes and workflows that, together, earn more money than it spends. If you can build a minimal version of the company that (1) can earn more money than it spends in the long run and (2) can be scaled up, you’ve answered many of the big questions of a venture. From there, your job becomes to make this engine more efficient, and grow it in size.

My favorite way to think about an MVP, and the way I believe most great founders approach prototyping, is to avoid thinking of “building” anything at all from the start. Instead, your goal as the founder is simply to come up with the most pressing questions, and get the highest-quality answers you can to those questions. Often, getting these questions answered will involve building significant parts of a business. But only build when a question necessitates it. This is what I call building for understanding.

The best founders I know build MVPs that are simply shortcuts and experiments to understand their customers better. To them, the fact that they are sometimes viable, working products is secondary. If you’re a founder working on a way to improve SMS marketing for small businesses, building a few small bespoke projects for small business will give you a far greater learning-to-cost ratio than building a full-fledged SaaS product, no matter how hacky. If you build a social app, researching which features to steal from the big social apps is probably the least productive thing you could do. Instead, study how specific niches of people spend time on their phones and computers. What are they doing more or less over time? What can you understand about these people better than other companies? If you have a specific feature in mind, put it in people’s hands and ask why questions – don’t simply measure numbers.

Conversely, if you’re an early stage founder or team before product-market fit, make sure you have very good reasons for doing anything that doesn’t improve your understanding of your customers. Building features for sake of “feature parity” against competitors might seem productive, but doesn’t help you learn anything new. If competitors are winning with a specific feature, they’ve already put in the time and money to validate its value. Your time is probably better spent investigating other features or problems, unless missing that specific feature is stopping you from answering some other important question (like, for example, “would people pay $X for what we have?").

Furthermore, not all knowledge-gathering is useful. A popular move for early founders is to run lots of surveys. Surveys are sometimes useful, but the vast majority of surveys done by first-time founders in the early stages of an idea don’t return data worth the effort put into gathering it. Surveys may get you some answers to yes or no questions, but it rarely helps you understand why people behave a certain way or want certain things. Making a survey is easy, but creating surveys that really help you understand people is tough, and a skill unto itself. In the beginning, your time is better served speaking to people and hearing their opinions by solving their problems firsthand.

The great thing about this understanding-focused view of an MVP is that it’s very difficult to lie to yourself about whether you understand something better. If you’re simply measuring numbers to see your progress, you can easily fudge the criteria here and there to make the numbers work. But if your measure of progress is “do I understand these people better than I did a week ago?” I usually find that the answer is pretty black and white.

Whenever I get a chance to talk to founders at the earliest stage, especially if they work in a field that interests me, I try to get a glimpse into what they’re learning now. The ones I admire the most always leave me with a new way to understand something, even if I knew it previously from a different angle.

Move fast to constantly sharpen your understanding, sometimes by building something new, other times by simply asking the right questions. That’s the best way I’ve found to start building something people want.


On Dorm Room Fund

Everything as processes

I share new posts on my newsletter. If you liked this one, you should consider joining the list.

Have a comment or response? You can email me.