Start Small, Build Big

Don't boil the ocean to make a cup of tea—launch tiny experiments and build from what works

Dhanji R. Prasanna
How Block is becoming the most AI-native enterprise in the world

Start Small, Build Big

"Start small with everything. If you're trying to boil the ocean to make a cup of tea... you'll never get there. So if you're making a cup of tea, just make the cup of tea. You don't need to boil all the water that there is." - Dhanji R. Prasanna

What It Is

Start Small, Build Big is Block's core operating tenet—and Dhanji's key leadership lesson. Rather than launching ambitious initiatives with large teams and comprehensive plans, start with the smallest possible experiment: one engineer, one idea, one weekend. If it works, build from there. If not, you've lost almost nothing.

This philosophy has driven Block's most successful products: Cash App started as a hack week idea, their Bitcoin product emerged from a three-person hackathon (including Jack Dorsey), and Goose began as one engineer's side project.

How It Works

The contrast:

Big launch (usually fails):

  • Google Wave: 70-80 engineers building before any real users
  • "Trying to be everything to everyone"
  • "Probably lacked some of that meeting the earth where reality lies and adapting accordingly"

Start small (Block's pattern):

  • Cash App: "Started more or less as a hack week sort of idea"
  • Bitcoin feature: Three people (Jack, Dhanji, one engineer) at a hackathon
  • Goose: One engineer building something useful, sharing with Databricks and Anthropic

Why small beats big:

  1. Faster feedback - Small experiments hit reality quickly
  2. Lower stakes - Failed experiments cost almost nothing
  3. Natural selection - Good ideas build momentum; bad ideas die quietly
  4. Authentic product-market fit - Products that survive the small stage have proven demand

The Carl Sagan corollary:

"If you want to make an apple pie from scratch, you have to first invent the universe."

The point: narrow your scope to what's achievable. You can't solve the whole problem at once, so don't try.

How to Apply It

  1. Question the scope:

    • When presented with a large project, ask: "What's the smallest version that would prove the concept?"
    • Could one engineer build a prototype in a week?
    • Could a hackathon team demonstrate value?

  2. Enable experiments:

    • Block has hack weeks, special projects (2-5 engineers each)
    • These create space for "starting small" without formal approval
    • Goose emerged from this kind of space

  3. Let momentum determine investment:

    • Don't staff up based on projected success
    • Start tiny, then add resources when something works
    • Cash App "grew into a bigger and bigger and bigger thing"

  4. Celebrate small beginnings:

    • Bitcoin feature: built in a hackathon, tested by buying coffee at Blue Bottle
    • Goose: one engineer's experiment that others got excited about
    • This creates culture where small experiments are respected

  5. Kill big proposals that skip small:

    • If a project requires 50 engineers from day one, it's probably wrong
    • Ask: "How would we test this with 3 people first?"

Examples from Block

Cash App:

  • Started as hack week idea
  • Grew organically as it worked
  • Now a major product with millions of users

Bitcoin feature:

  • Three-person hackathon project (Jack, Dhanji, one engineer)
  • First tested by buying coffee at Blue Bottle
  • Block became the first public company to launch Bitcoin product

Goose:

  • Brad (creator) built it on his own time
  • Shared it with Databricks and Anthropic to get feedback
  • Built momentum organically
  • Now used across Block and many other companies

When to Use It

  • At the start of any new initiative
  • When planning product expansions
  • When evaluating ambitious proposals
  • When resources are constrained (always)
  • When past big launches have failed

Source

  • Guest: Dhanji R. Prasanna
  • Episode: "How Block is becoming the most AI-native enterprise in the world"
  • Key Discussion: (01:09:19) - Start small principle with examples
  • YouTube: Watch on YouTube

Related Frameworks

Back to Execution