After six months of on-the-ground visits to Yiwu, Shuibei, and Mexico, a Web3 payments entrepreneur reflects deeply on why she decided to exit. She reveals that the real moat in payments is not the product but banking relationships, licenses, and risk control, and clarifies the true penetration rate and structural oppo
Author: Yuki (Liu Yuqing), Stablehunter/Money in Motion
Over the past six months, I went from being an outsider to Web3 to stepping inside the payments industry. And now, I've chosen to stop and no longer continue doing Web3 payments.
This is not a retreat after failure, but an adjustment of judgment made after truly getting into the field. Over these six months, I went to Yiwu, Shuibei, and Putian, and also to Mexico, to see how payments are actually made in the places that reports describe as the most bustling. I also got into the field myself—built an MVP for Web3 payments, took over accounts, made Web3 collection tools, and tried to run the imagined path from the first step all the way to the last.
But the deeper I went, the more clearly I realized one thing: this is not an industry where "making a good product means you win." Payments don't compete on features, but on banking relationships, licenses, capital efficiency, and the long-term ability to control risk.
Many payment businesses that look "profitable" are essentially earning not a capability premium but a risk premium—it's just that nothing has gone wrong yet. What truly determines how far a payments company can go has never been how much money it makes, but whether it can still withstand and survive before the risk actually materializes .
This article is not meant to deny the industry, but to remove the filter, lay out the real structure, and leave some clearer judgment for those who come after. (A few weeks ago, I also recorded a podcast with former Kun Global VP Robert, Nayuta Capital CEO, and former Didi Finance CEO Alex, discussing the same questions.)
As a serial entrepreneur, I ended a multi-year startup project last year. During the process of shutting down the company, I also set aside some time to rest, returning to a more "emptied-out" position to seriously think about what direction I should focus my energy on next.
Six months ago, a friend invited me to Hong Kong to try starting a Web3 payments venture together. At the time, I wasn't familiar with Web3 itself, and I couldn't say I had any real understanding of the payments industry. From a macro perspective alone, it was clearly an industry large enough and still in an upward cycle, and at the same time, there was potential room for Web3 and AI to combine.
In my previous entrepreneurial experience, we had done cross-border business and also built platforms and software related to remote work. In these practices, I kept running into the same fact: business can go global very quickly, but the flow of funds always lags behind. Slow settlement, fragmented paths, opaque costs, uncontrollable payment terms—these problems might be worked around through experience and patience when the scale is still small; but once the business scales up, they won't be solved by "management capability," they will only be continuously amplified. Money cannot flow as freely as information, and this itself is the invisible ceiling of many globalized businesses.
It was precisely against this backdrop that, when I began to systematically understand how Web3 payments are actually used at the clearing and settlement level, what it presented was not an abstract technical narrative, but a solution that could logically act directly on these pain points: faster settlement speed, higher transparency, and near-round-the-clock clearing capability.
In my judgment at the time, this looked like a direction that both solves real problems and is Day 1 Global —I didn't enter because of Web3 itself, but because in this specific payments scenario, it seemed to offer a better structure—at least logically, it seemed enough to pry open those frictions that had long existed but been ignored.
But looking back now, I gradually realized that at the time, like many people, I had defaulted to a premise that reality kept challenging: as long as clearing and settlement efficiency is high enough, payments will naturally migrate on-chain. It was even further simplified into an intuition—payments are just matching transactions, and as long as you get the process running, you can "hand-craft" cash flow.
Based on my lack of understanding of Web3 and the payments industry, I decided at the time to first spend three months truly getting into this industry, figuring out the structure, and then deciding what exactly to do and what position to do it from.
When I arrived in Hong Kong, the initial idea wasn't complicated. The original thought was very simple: relying on some resources and relationships my friend already had, start from OTC or relatively simple collection and payment scenarios, get the cash flow running first, and then decide what to do next based on real demand.
I wasn't there to do research, nor to watch from the sidelines long-term, but to see—whether it was possible to first build something that could run, and then calibrate the direction in real business.
But very quickly, the external environment underwent a clear acceleration. In May, the U.S. passed the GENIUS Act, and the entire industry was ignited almost overnight. Capital, projects, and entrepreneurs poured in rapidly, and Web3 payments went from a relatively niche infrastructure topic to a frequently discussed "new opportunity." From the outside, this was a positive; but for a startup team that had just entered the field, this sudden bustle was actually not a good thing.
The more mixed, noisy, and quickly consensus-forming a moment is, the more easily it conceals the real problems. Internet giants, financial institutions, banks, traditional Web2 payment companies, and Web3 Native teams entered one after another. Everyone was talking about opportunities, but few were talking about structure. And I felt at the time that I should sink even more into the front lines and truly figure out this industry.
After I truly started working the front lines, the first thing I did was not to continue optimizing the product plan, but to look at: who is actually using Web3 payments? Why? Where? I first went to Yiwu, the place most frequently mentioned in reports.
In many studies and shares, Yiwu is often used as a representative sample of "Web3 payments already being applied at scale." But when I actually walked through it, what I saw was a different picture. Stablecoins do exist, but more as scattered, relationship-driven, hidden usage .
It has not become a settlement method that can be standardized and productized for replication, as described in reports. Many transactions are not because of "optimal efficiency." After that, I went to Shuibei, Putian, and Mexico, and also learned about the penetration rates in Africa, Argentina, and other places—the situation was not fundamentally different.
Web3 payments are not nonexistent, but they are far from having formed a stable, scalable main path; more often they are just a "patch" embedded in the existing system. The real penetration rate does not match the heat we perceive in reports, communities, and discussions.
But it was precisely in these exchanges that I gradually shifted my perspective from "can I make a product" to the industry structure itself . I began to realize that the incremental market for stablecoins is probably not "inside the crypto circle," but in those business scenarios in the Web2 world that already exist but have long been slowed down by the traditional clearing and settlement system.
This is not a narrative shift, but more like a slowly unfolding fintech upgrade. At the same time, a question began to emerge: if real usage is so fragmented, does the productization path actually hold up?
From July to September, I continued field research while systematically reaching out to potential customers. HR companies, insurance, tourism, MCN agencies, service trade, cross-border businesses, gaming companies... the needs varied, but the core problem they pointed to was highly consistent: money should flow faster, cheaper, and more stably.
Payroll, t