Forty-five apps of practice
Benjamin Chen is twenty years old, studies at NYU, and has built more than forty-five apps. That number is the whole story in miniature. Most founders get a handful of launches in a career. Chen turned launching into a practice routine, and Snag — his app for finding free items listed in your local area — is what the routine produced at full maturity.
Snag's premise is almost insultingly simple. People give away free stuff constantly: furniture, electronics, clothes, moving boxes. The listings exist scattered across marketplaces and curb alerts. Snag aggregates them into one feed organized by proximity, so users can claim free items near them before anyone else does.
The app scaled to $30,000 in monthly recurring revenue in just four months. For context, that is a pace most venture-backed consumer apps would celebrate, achieved by a college student operating a personal system refined across dozens of prior attempts.
What separates Chen from the average prolific builder is that he treats the process itself as the product. Every app is an experiment run through the same four-step playbook, and each run sharpens the playbook for the next. Snag is not a lucky hit. It is the output of a machine he spent years building.
Start with an idea that sells itself
Step one of Chen's system is idea selection, and his criterion is specific: find a scalable idea anchored by a simple, powerful value proposition. The best candidates directly help users make or save money, because that kind of value converts almost automatically.
Snag is the textbook case. Free stuff near you needs no explanation, no education, no trust-building. The value is immediate and tangible the first time you open the app. That built-in conversion rate is doing enormously quiet work in the background of every growth metric — acquisition, activation, retention, and word of mouth all benefit when the proposition sells itself.
This discipline matters more than it looks. Most app ideas require the founder to manufacture desire: convince users they have a problem, then convince them your solution fits. Chen skips both steps by picking ideas where the desire is pre-installed. Nobody needs to be persuaded that free furniture nearby is worth a tap.
The filter also keeps him honest about scalability. An idea that converts well but cannot reach millions of users is a lifestyle business, not a system output. Snag's market — everyone who likes free things, which is everyone — clears that bar comfortably.
Ship the MVP in an afternoon
Step two is speed. Chen builds minimum viable products in hours, not weeks, using AI programming tools — specifically Claude Code running inside the Cursor IDE. He has said he consistently ships functional apps with the core build done in four to five hours.
That number should recalibrate every founder's sense of what v1 requires. Four hours does not produce a polished product. It produces something testable, which is all an MVP needs to be. Polish comes later, and only for the ideas that earn it through traction.
The speed compounds with the volume. When each attempt costs an afternoon rather than a quarter, you can run forty-five experiments in the time most founders run three. Chen's hit rate does not need to be high. It needs to be nonzero across enough shots, and the system guarantees the shots keep coming.
There is a subtler advantage too. Building fast keeps emotional attachment low. When you spent four hours on something, killing it after a week of flat metrics is easy. When you spent four months, you negotiate with the data. Chen's system is designed to make quitting cheap, which is what makes persisting with winners possible.
Distribution and iteration close the loop
Steps three and four are where apps usually die, and where Chen's system is most opinionated. Step three is distribution, handled with the same scrappiness as the build: Chen films marketing videos himself and systematically tests user-generated content to find what pulls.
Founder-shot video has structural advantages. It is free, so testing volume stays high. It is authentic in a way polished ads are not, which matters for an app whose whole pitch is real free stuff from real neighbors. And the feedback loop is immediate — post in the morning, read the metrics at night, reshoot tomorrow.
Step four is relentless iteration driven by data. Chen uses Mixpanel analytics and direct user feedback to understand behavior and refine the experience continuously. Every cohort teaches something: which onboarding step leaks, which notification brings people back, which category of free items drives the most sessions. The compounding goal is explicit — push lifetime value up and customer acquisition cost down, simultaneously.
Put together, the four steps form a flywheel. A converting idea makes distribution cheaper. Fast builds make more ideas testable. Self-shot marketing makes feedback instant. Analytics make iteration precise. Run it forty-five times and eventually you get Snag: $30K a month in four months, from a system, not a stroke of luck.
The playbook
- Pick ideas with built-in conversion — Snag helps users find free stuff nearby — a value proposition with instant, tangible payoff. Chen picks ideas where the benefit is obvious in one sentence, because obvious converts.
- Build the MVP in hours, not weeks — Core builds ship in four to five hours using Claude Code inside Cursor. Speed is a strategy: the faster you ship, the more shots you get.
- Do your own distribution — Chen films marketing videos himself and systematically tests UGC. Founder-shot content is cheap, authentic, and iterates as fast as the product.
- Let analytics dictate the roadmap — Mixpanel analytics and user feedback drive every iteration. Watch what users actually do, then refine relentlessly toward higher LTV and lower CAC.
- Make it a system, not a lottery ticket — The system matters more than any single app. 45 apps of practice turned launching into a repeatable process instead of a gamble.