Junxian Ma / Writings

Every conversation with outstanding entrepreneurs brings a very distinct kind of new insight. In the days leading up to N1, I have met more and more like-minded people. Talking with them has given me a few new thoughts.

The first is to iterate faster. In the past, I often subconsciously wanted to think things through more completely before starting. I wanted to prove that a direction was sufficiently correct before taking action, or at least avoid making obvious mistakes. But recently I have increasingly felt that many judgments are not produced by static reasoning in my head. They grow out of repeated contact with real feedback. If an idea stays only as an idea, it can easily look reasonable, beautiful, and complete. Only when it becomes something other people can actually see, use, and evaluate does it begin to reveal its real problems.

So one important thing for the next stage is to make my unit of action smaller: to turn “think clearly before doing” into “do a little, then understand it a little more clearly.” This does not mean giving up thinking. It means not allowing thinking to become an excuse for delaying action. A prototype, an interview, a landing page, a demo, or one piece of feedback from a real user may be more useful than thinking alone for many days. Faster iteration means touching reality faster. It also means admitting more quickly where I am wrong.

The second is to perceive problems more sharply. Many truly important problems do not first appear in the form of “startup opportunities.” At the beginning, they may simply be something someone repeatedly complains about, a small friction in a workflow, or a part of an industry that everyone has become used to even though it is clearly inefficient. Excellent entrepreneurs seem to have a certain ability: they can detect unreasonable things in very ordinary daily experience that others overlook.

I do not think this kind of perception comes from nowhere. It comes from continuously putting oneself in real environments, talking to real people, and observing how they work, make decisions, and get stuck. Simply asking “What do you need?” is often not enough, because many people do not express their real problems in the language of needs. What matters more is observing what they actually do: where they pause, where they take detours, and where they have already built an ugly temporary workaround. Those places may be where the real problem begins to shine.

The third is to conduct broader research. By research, I do not only mean reading reports, checking competitors, or reading papers. I mean placing a problem inside a larger system: Why is it appearing now? Who feels the pain most strongly? How are they solving it today? Why have existing solutions not fully solved it? Are there changes in technology, market structure, organizations, or culture behind it? If I only stare at one point, it is easy to mistake local excitement for a long-term opportunity. But if I look broadly enough, it becomes easier to judge whether a problem is worth committing to for a long time.

I also realize that research is not meant to make me more conservative. It is meant to make me take risks more accurately. Even after research, I may still be wrong. But at least I will know what I am betting on, and which assumptions need to be tested most urgently. Good research should serve action, not replace it. In the end, it should produce a clearer next experiment, not a pile of information that looks rich but does not change any decision.

The fourth is to think more deeply. Fast iteration and broad research are important, but without deep thinking, it is easy to become someone who merely chases surface-level feedback. Feedback itself does not automatically tell me the direction. What users say is not always the essence of the problem. The difficult part is how to abstract structure from scattered information, how to judge which phenomena are noise and which represent deeper trends, and how to connect short-term actions with long-term goals.

This also makes me understand “strategy” differently. It may not mean making a grand plan at the very beginning. Instead, through repeated action and feedback, strategy may gradually become a clearer set of judgment criteria: what problems are worth solving, what feedback is worth listening to, what opportunities should be abandoned, and what things are worth persisting in even when short-term results are unclear. Deep thinking is not slowness. It is the ability to avoid being pulled around by surface phenomena in a rapidly changing environment.

Putting these feelings together, I think the most important state for the next stage is to act fast enough, observe sharply enough, take in input broadly enough, and reflect honestly enough. I should not only be someone who thinks a lot, nor only someone who moves quickly. I hope to build a loop between the two: action brings feedback, feedback drives research, research settles into judgment, and judgment guides the next round of action.

So after N1 begins, I hope to at least do a few things:

• Continue having deep conversations with people from different backgrounds every week, not only to hear their stories, but to understand the real problems they face.
• Turn ideas into things that others can evaluate as soon as possible, no matter how rough they are at the beginning.
• Record and review every piece of feedback, not only the conclusion, but also the reasoning behind my judgment at the time.
• Maintain enough depth in one technical direction, so that I do not remain only at the level of vague trend discussions.
• Write logs regularly, recording my current confusions, judgments, and actions, so that my thinking can be revisited and corrected.

I still do not know what I will ultimately do, and I am not sure which path will really lead to the larger goal. But at least at this stage, I am beginning to understand more clearly that direction is not something to wait for. It is something calibrated through continuous action. I hope I can move a little faster, and stay a little clearer.