Start Small, Don't Boil the Ocean
Narrow scope to the achievable thing in front of you, then build momentum on top of it
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 3
- Confidence
- 85%
Prasanna's core operating tenet: 'if you try to boil the ocean to make a cup of tea, you'll never get there.' Rather than resourcing a big idea heavily on day one, narrow scope to the small achievable thing, ship it, and let it accrue momentum. Nearly all of Block's biggest wins — Goose, Cash App, its Bitcoin product — began as tiny experiments; its biggest failures began by going big before finding reality.
Origin
Prasanna's long-standing tenet at Block/Square; he cites the aphorism 'to make an apple pie from scratch you must first invent the universe' (attributed to Carl Sagan) as a companion idea.
Core principles
- 01Narrow scope to the thing in front of you that's actually achievable
- 02Momentum compounds: small working thing → build on top → bigger thing
- 03Going big on day one, before real users, is a failure pattern
- 04Small experiments that satisfy a clear thesis are the seed of large businesses
- 05A working useful product for real people beats a grand vision with no users
How to run it
- 1
Shrink the goal to a cup of tea
Reframe the big ambition down to the single smallest achievable version. Don't provision resources for the whole ocean.
Pro tip Make the actual cup of tea — don't boil all the water there is.
- 2
Ship a proof of concept from a thesis
Have someone build a small working artifact that satisfies a clear belief, and share it to generate excitement and learning.
Pro tip Share the POC with external partners early to learn and build momentum (Goose was shared with Databricks and Anthropic).
- 3
Build on top as momentum grows
Let the small win accrue interest and expand it iteratively into the bigger thing, rather than launching big.
Pro tip Hack-week-scale ideas can become giant businesses — protect the path from experiment to product.
Watch out Starting big with 70–80 engineers before you have users (Google Wave) lacks the reality-check that scoping small provides.
In the wild
Goose started as Block engineer Brad building on his own time from a thesis — long before the buzzword — that agents would be how LLMs unlock value. He built a proof of concept and shared it with Databricks and Anthropic, who got excited, and momentum grew from there.
→ Became Block's widely-used open-source general-purpose AI agent, adopted internally and by competitors and partners including Databricks.
At Google, Wave 'tried to be everything to everyone' with 70–80 engineers building it before it had any users outside Google. It launched big on day one.
→ An 'epic failure' — it lacked the reality-grounding that starting small provides; Prasanna extracted the lesson that going big before finding users is a failure pattern.
Common mistakes
Resourcing a big idea heavily on day one
Going big before real users (Google Wave, 70–80 engineers) skips the reality check; you build the wrong thing at scale.
Boiling the ocean to make a cup of tea
Trying to build the complete, universal version blocks you from ever shipping the small achievable thing that would actually reach users.
Is it for you?
Best for
Founders and intrapreneurs tempted to fully resource an ambitious vision before validating a small core
Not ideal for
Efforts that genuinely require large upfront coordinated investment (heavy infrastructure, regulated hardware) where a toy version can't validate the thesis
From the transcript
“if you try to boil the ocean to make a cup of tea, I don't know who said that, but it's a really a useful…”
“narrow your scope to the thing that's in front of you and that's achievable”
“Goose started small. It was just an engineer working on their own time”
“we were 70 80 engineers building this thing before it even really had any users outside Google”
From the episode
How Block is becoming the most AI-native enterprise in the world
Dhanji R. Prasanna