Zero to One Means Pruning Fast

Pruning Beats Building Speed

Most startup advice celebrates shipping velocity. Casey Winters, writing about building SuperMe, points somewhere else.[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

His closing line: "Zero to one isn't about building fast. It's about pruning fast."[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

Before product/market fit, you will try many things that do not work. The deciding skill is spotting those failures and removing them quickly.

Shipping more features than competitors mostly gives you more things to judge.

Surface Area and Conviction

Winters says building SuperMe made something obvious: zero to one is about surface area and conviction.[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

Surface area is how much a team builds. Every feature, flow, and experiment adds to it.

Conviction is knowing what the company must own. Winters ties every cut back to it.[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

More surface area gives you more to learn from, and more to cut. Conviction tells you which parts deserve protection.

Know What You Must Own

Winters warns: "If you don't know what you need to own, you won't know what to remove."[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

That turns conviction into a working filter.

For each piece of surface area, ask two questions. Is it part of what we must own? Is it working?

Core and working: build on it. Core but weak: polish it. Not part of what you must own: it is a candidate to cut.

What Pruning Is Not

Pruning fast does not mean trying less. The essay assumes you will try things that do not work.[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

Cutting without knowing what you must own is not pruning either. It is just deleting.

Real pruning needs both: plenty of experiments, and a clear idea of what to protect.

Same Founders, Different Contexts

Winters and co-founder Ludo first saw each other's styles while scaling Pinterest.[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

Ludo had an intuition for when to hack and when to scale engineering systems. Winters could strategically prioritize and frame problems for the rest of the organization.[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

At Whatnot, Ludo operated at more scale. Winters, a part-time advisor, was "more direct and less patient."[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

Pinterest and Whatnot already had product/market fit. SuperMe started from scratch.[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

Why Old Partners Can Differ

After years together, Winters expected to know Ludo's style. Yet their approaches to zero to one turned out "quite different."[[1]](https://www.caseyaccidental.com/p/zero-to-one-subtraction)

The reason is context. Scaling, advising part-time, and building from scratch reward different behaviors.

An operating style is not a fixed trait. What worked while scaling may not transfer to a blank page.

So talk early about how you will prune, even with a partner you trust.

Your Pruning Playbook

Here is how to use Winters's idea on your own early product.

Expect experiments to fail. Those failures are the raw material for pruning.

Measure your speed by how quickly you remove what is not working, rather than by how much ships.

And agree with co-founders on what you must own before you start cutting.

Before product/market fit, is recognizing and removing failed ideas usually more important than shipping more features quickly?

• Yes
• No

Answer: Yes

Yes. Early teams will try many things that fail, so progress depends on spotting and removing failures quickly. Shipping speed alone can create more features to evaluate without improving focus.

A pruning cycle starts by _____, then _____, then _____, and finally _____.

Answer: A pruning cycle starts by trying things, then naming the core, then reviewing features, and finally cutting on purpose.

The cycle first creates enough surface area to learn, identifies what the company must own, reviews each feature, and removes what is not working. Partner alignment is also important, but these four steps capture the core pruning sequence.

A team reviews a feature that is central to what the company must own but is performing weakly. What should the team do?

• Build on it immediately
• Polish it
• Cut it because it is weak
• Keep it without changing anything

Answer: Polish it

A core feature that is not working should be polished. The team should protect it because it is central to the company’s conviction, while improving its performance.