portotype.com

Running a Startup - Three Layers theory

Someone asked me how I run my Board. Well, I think your Board meetings should mimic how you run your company.

So here is how I like to run a high potential tech company. It starts with 3 key ingredients [1]:

  1. The team and investors believe in a Mission and embody it;
  2. The team is accountable to an ambitious Process of Iterations;
  3. You Focus on Customer Value as the results that fuel more progress.

Founders, Team, Investors should be on board with all three, all of the time.

Yes, I think thay every big decision, every hire, every planning session, every project must exist in the universe where those 3 layers exist.

Otherwise, your business will be built on a slightly more tactical way, and you will be wasting the potential offered by ever faster, every changing, ever more powerful technology.

True Layers What your business looks like
just Results a manager working for a business with a bank loan
just Process a consultant
just Mission vaporware and hype based investment
just Results + Process an expert working in a PE-backed company
Results + Process + Mission an entrepreneur backed by a VC

As an example, lets take a look at Board meetings.

Your Board meetings should always strongly feature these 3 sections very clear: Vision, Iterations (plan) and Results.

Never let any of the layers die out!


  1. The Vision is an aspirational objective that is hard to get to, likely impossible to ever make perfect, like i) to connect everyone in the world or ii) to solve pancreatic cancer, or iii) to give computation to the 99% of non-engineers, or iv) to build the classiest, lightest, and most maintenance-free bicyle in the world.
  2. Iterations are the experiments you will run to get there. There should be projects to double down on things that are working well, and others to fix what is broken. Some will be iterative and look at the next months, and others should push the barrier for years to come. Examples: i) experiment different ad copy to double paid acquisition volume at same price (pretty lame experiment), or ii) create a vertical full-stack markdown as a way to encode business logic for industry X in a human-ready format that AI can then convert to code, or iii) let users remix code in a way that is like repository cloning but simpler and more fun, to get each user to bring 1.3 new users, or iv) try different aluminum, carbon and titanium alloys to achieve a timeless, light, sturdy frame.
  3. Results are what makes a business last, customer satisfaction, purchases. revenue.