top of page

How to start


Orange, Red, Yellow by Mark Rothko, 1961


Life should be really simple but we insist on making it complicated - Confucius


Startups shouldn’t be complicated either. Don’t get me wrong, I know building startups is hard, very hard. That’s precisely the reason entrepreneurs should start with the simplest of plans.


Minimalism is the best antidote to the messy nature of entrepreneurial innovation. Entrepreneurs should keep the starting up process minimal: a plan reduced to a stage where no one can remove anything further to improve its design.


I submit the simple rule to starting startups.


To start, entrepreneurs should target a small market and solve a single problem with a simple value proposition.


Start with a small market


Targeting a small market allows founders to focus on a homogeneous group with homogeneous needs. Such markets can be found in a single neighborhood or a digital community.


For example, when Brex started providing credit cards to startups, it focused on serving its YC batch colleagues first, then the YC community before expanding to startups in general.


The smaller the starting market the easier it is to find, serve and retain the target market. In Zero to One, Peter Thiel and Blake Masters argue that startups should adopt a monopoly approach from the start. A perfect target market should be a small group of people concentrated together and served by few or no competitor.


Before Cornershop scaled to many countries, it established a small monopoly in Mexico City and Monterrey by serving a handful of upper middle class neighborhoods. At some point, it had more than 80% of on demand groceries market share. From that tiny market it built one of the most successful marketplaces in the region.


We see many founders focus on very large heterogeneous markets or diversify to multiple small markets to ‘de-risk’ the inquiry process or to please investors with a big vision. In our experience it makes the difficult mission of finding a product-market even more difficult.


It’s hard to build great software for pharmacies, gift shops and spas simultaneously. It’s hard to find the right marketplace offering for all members of the Colombian middle class.


We love small beachhead markets.


Start by targeting the tour operators in Cancun or by serving young middle class moms in Santiago.


What’s important for us is that you envision a huge addressable market over time but start as small as possible.


Once you find a product market fit you can expand and go big: go for all tour operators in the world. Go for all demographics that value your product or service as much as the early adopters.


Focus on a single problem


In Four Steps to the Epiphany, Steve Blank argues that startups should focus on customer development rather than product development to emphasize learning about their problems as early as possible. I agree.


In his 4-step customer development methodology, the first two steps are all about honing in on a single problem: customer discovery and customer validation. First, test hypotheses about the nature of the problem and gauge interest in the solution. Then validate the willingness for customers to pay and design possible distribution channels.


Back to our Cancun tour operator, it’s easy to imagine all the problems it has. It needs marketing to generate leads, it needs a way to manage reservations and payments. It needs some kind of accounting to manage their business and some way to pay and manage its employees. It needs credit. Businesses always need credit, don’t they?


The typical founder might present a slide with all the problems that technology could solve for tour operators to make their life easier and grow their business. Founders might not want to decide where to start so they build for several problems before getting the first user.


Solving a single problem means choosing where to start. Do tour operators care about their accounting and payroll management, maybe. Does it keep them up at night, I doubt it.


Focus on your customers’ dreams and nightmares.


Same goes for young moms. You can sell them everything they buy under one digital roof for young moms. Will your tiny startup be good at selling everything young moms buy, hardly, not at the start for sure.


Entrepreneurs almost always are too ambitious for what they want to build in year 1 and not ambitious enough with year 10.


Amazon started selling books. The rise of the internet allowed Jeff Bezos to offer 100x more books than a typical Barnes & Noble and deliver them directly to customers' front doors. He couldn’t possibly conquer all retail online, had he tried to sell everything day 1.


Design a simple value proposition


Once you decide to focus on a single problem, you can focus on making your solution as simple and intuitive as possible.


When Eric Reis introduced the MVP concept in Lean Startup, he simply described the way startups were built by the best tech entrepreneurs. More than its ‘size’, MVPs are a tool to test entrepreneurial hypotheses.


Reis describes it as a "version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort"


At its core, it’s the simplest viable value proposition.


A good value proposition should feel obvious when explained to customers. You almost want to hear a sigh of relief or a wow of curious skepticism when your potential customers see the demo.


Stripe’s simplicity was key to its value proposition.


Add 7 lines of code and get access to a powerful payment processor API. The developers, an important decision maker within any potential customer, appreciated the elegance of the solution. A big step forward from nightmarish incumbent payment systems.

Simple value propositions minimize any noise in customer’s feedback. There’s no need to validate if the sale’s team properly explains it or if the copy on the landing page properly conveys the solution.

When Fintual launched its stock trading platform in Chile, it conveyed its value proposition in the simplest terms: Own a piece of this company. Invest as little as $1 in stocks and ETFs.


While the MVP concept has been widely adopted, founders keep making their first product way too complicated with way too many bells and whistles. Customers don’t want an all-in-one-something or a one-stop-whatever or a super-duper-app.


Customers need a product that does one job, well and every time.


Brian Chesky, founder of Airbnb argues that your company needs to deserve expanding to a new feature or vertical. It deserves it when customers are entirely satisfied with the first product.


Testing one hypothesis at a time


Before finding the elusive product-market fit, entrepreneurs will probably need to test several hypotheses.


For Reis, a pivot is a "structured course correction designed to test a new fundamental hypothesis about the product, strategy, and engine of growth."


You should see the simple startup maxim as a three legged stool, pivoting means you switch a single element of your hypothesis. One leg at a time.


The inquiry process may reveal a new more important problem. Founders might discover more competition than expected in the first beachhead market. The first value proposition may get a lukewarm response. Pivot, build, repeat.


At the heart of starting well are focus and speed. Always keep it simple, test one hypothesis at a time, and build fast.


Every legendary company started small.










231 views

SUBSCRIBE to keep up with my writing. 

I write about my work as an investor, a lecturer, and a mentor. In general, musings about Latin American tech, VC and life.

bottom of page