At the beginning of a new project, everything seems to be on the table and nothing truly exists.
There are many opinions, scattered user voices, technology that can be built quickly and a goal that sounds very large. People discuss product form, business models and long-term imagination, but may still not know what should happen tomorrow.
This is usually the stage that gives me the most energy.
I like picking up the scattered pieces and finding the shortest path: begin with a real problem, make the smallest experience, find the first users and let the result decide the next step.
The first path is not the first complete product
When we built Mengya, we did not begin by developing a fully featured application.
I first turned the problem into a mockup and used Xiaohongshu to recruit external users while moving a minimal conversation experience forward. We wrote the conditions for continuing and stopping in advance: would people apply, would they actually speak, would the conversation create value, and would they return the next day?
In the end, 62 users outside Fanshu joined the recruitment. Seventy-two people registered, 49 began a conversation and 18 continued for more than ten turns. Acquisition and first value worked. Next-day retention was only 15%, below the 30% target.
The path did not lead to a successful product, but it made the uncertain problem concrete. People were willing to arrive and could feel surprised by being “organized” and “seen,” but the current form gave them no lasting reason to return.
If we had built a large first version, we might have tried to explain the result with more features. Because the path was short, the answer was clearer: the original assumption did not hold, and we should stop.
Zero to one is not a story about one person doing everything
From the outside, zero-to-one work can become a heroic narrative: one person invents the idea, draws the screens, writes the copy, finds users, runs operations and finishes everything.
I do often work across role boundaries. Apart from technical implementation, I try to connect the user journey, product structure, PRD, interface, recruitment, interviews, business model and review. I do this not because every task must belong to me, but because a new project is most vulnerable when the important questions are split across roles and no one connects them into a decision loop.
“I did everything” is not the capability. The real capability is being able to state:
- what this experiment is testing;
- what evidence is enough to continue;
- which result should make us stop;
- and how someone else can take over once the path works.
An engineer owning implementation does not reduce my responsibility for the overall path. In the same way, my responsibility for the path does not allow me to claim another person’s technical work. Zero to one needs collaboration with clear boundaries, not blurred contribution.
The first path ends when the project no longer lives only in one person’s head
Some people are excellent at making an established business more stable: refining processes, managing teams, improving conversion and optimizing the same system day after day. That is a difficult capability in its own right.
I am more drawn to the moment before the answer exists. I want to find real users and compress a vague problem into something testable. When the path works, I can hand the process, measures and key judgments to people who are better at scaling it. When it does not, the work should still leave an honest review rather than hiding the failure.
Moving fast does not mean making endless new things. Experimentation without stopping conditions creates a pile of half-products. Pathfinding without records and handover only leaves disorder for the people who follow.
A path that has truly been opened should tell the team where it began, what it passed through, where people are likely to get lost and how the next attempt can become steadier.
I like zero-to-one work not because a blank page feels free.
I like it because making the first path clear in the middle of confusion allows everyone to continue from the same reality.