45 Apps in a Year, Then Snag — How a 20-Year-Old NYU Student Turned a “Free Stuff Near You” App into $30K/Month in Under 4 Months

Snag helps you find free stuff nearby — furniture, electronics, everyday goods — in one place. It scans across online boards, local groups, marketplaces and community feeds, and surfaces only the items being given away for $0. Users pay a few dollars a month to access those free finds they’d otherwise have to hunt down across a dozen sources.

It was built by Benjamin Chen — ‘Benji’ — a 20-year-old NYU student. The notable part: Snag isn’t his first hit. Over the past year he shipped 45+ apps, and through that repetition he sharpened a repeatable system for how to build and distribute a consumer app. Snag rode that system to ~$30,000/month, 9,000 paying conversions and 100,000+ authenticated users in under four months, listed under Shopping with ~3,300 App Store ratings.

Technically it’s optimized for shipping fast, not for craft. Coding is done with Claude Code, the backend on Supabase, design in Figma — a standardized line that turns out an app (minus backend) in ~4–5 hours (with Loops for email and Mixpanel for analytics). The products are deliberately single-purpose — thin apps around one clear use case — produced in volume. The edge isn’t inside the app; it’s in what happens after it ships.

The growth engine is two-stage: recruit many creators to make UGC-style videos, then promote only the organically-winning videos as Meta (Facebook/Instagram) performance ads. Pricing spans weekly/monthly/yearly/one-time but stays at ‘a few dollars a month’ to maximize conversion. In short, Snag’s story isn’t about a genius app — it’s about a factory that mass-produces apps and drops them onto a standardized distribution pipeline.

The repeatable playbook

1. 1Don’t bet on one app; drop build cost to 4–5 hours so each app is a trial, not a bet (ship single-purpose apps in volume)
2. 2Ship fast on a standard line: design in Figma → generate with Claude Code → backend on Supabase (bake in auth to clear review)
3. 3Screen creators by interview for ‘virality built in’ (~10% pass), then retainer + CPM to mass-produce videos
4. 4Before spending on ads, make ‘50K organic views’ the qualifying bar for a winning video (select on free reach)
5. 5Reuse winners as Meta ads: $50 test → scale only ROAS>1 to $100 → $200 → $300 (refill fatigued ads with new winners)
6. 6Price as an asymmetric ‘a few dollars for hundreds of dollars’ offer and recoup on conversion (adoption over ARPU)

Deep dive

【Deep dive】Snag’s real product isn’t ‘the Snag app.’ It’s the repeatable system: build in 4–5 hours, drop onto a standardized pipeline, amplify only the winners with ads. Here’s the factory line, taken apart.

■ The product is ‘disposable,’ the asset is the ‘factory’ — why 45 apps. Benji shipped 45+ apps in a year. Most fizzle, but that’s by design, not failure. Drive the cost of one app down to 4–5 hours and an app stops being a bet and becomes a trial. Across 45 attempts, meta-level lessons accumulate — which single-purpose ideas land, which video angles spread — and those lessons raise the odds of drawing a winner like Snag. The mindset shift for indie builders: don’t bet on one app; ship many, cheap and fast, and let the system find the hit.

■ Build in 4–5 hours: the Claude Code + Supabase + Figma line. Benji’s build follows the same shape every time — reverse-engineer the value proposition, wireframe and design the UI in Figma, hand the designs to Claude Code in the IDE to generate the code, host the backend on Supabase. Because each app is a thin single-purpose wrapper around one API, this takes ~4–5 hours minus the backend. Crucially, account creation and auth are baked in from the start so the app clears App Store review (skimp here and the production line jams in approval). He doesn’t win on technology; he reduces technology to a fast, repeatable routine.

■ Growth engine, part 1: the creator-sourcing funnel (interview 100, keep 10). Half of Snag’s growth is decided at ‘hiring.’ Benji reaches out to many creators and interviews for ‘virality built in.’ The pass rate is ~10% — keep only 9–10 of every 100. The survivors go on a monthly retainer plus CPM and mass-produce videos. The gate that matters: only videos that clear 50,000 organic views advance to the ad stage. In other words, winning creatives are selected by free view counts before a dollar of ad spend.

■ Growth engine, part 2: validate organically, then move to Meta ads (the ROAS ladder). Winning videos (50K+ views) become Meta ad creative. He doesn’t go big immediately: a $50/day test measures ROAS and CTR, and only ads with ROAS above 1 / high CTR scale up $50 → $100 → $200 → $300. He watches for ad fatigue and keeps refilling with fresh winners from the creator pipeline. As a rule of thumb, every ~100,000 UGC views yields ~$1–2K in subscription profit. The point is that ‘prove demand, then spend’ is operationalized through creator supply and numeric gates.

■ Pricing & paywall: ‘a few dollars for hundreds of dollars’ — conversion via perceived value. Pricing spans weekly/monthly/yearly/one-time but stays at ‘a few dollars a month.’ The logic is clean: Snag promises access to hundreds of dollars’ worth of free goods, so a few dollars looks absurdly cheap against perceived value. That asymmetric offer drives a high conversion rate (9,000 paying out of 100K users). Acquire cheap and wide, recoup on conversion — the Cal-AI-style ‘adoption over ARPU,’ supercharged by the strong pull of free finds.

■ The limits of repeatability: this is a hit-rate game. Honestly, this playbook won’t land anyone’s first app. Snag shines because 45 prior iterations, an established creator supply network and a dialed-in ad operation came first. And the economics lean hard on UGC-to-Meta ROAS; when ads fatigue, the numbers soften. An aggregator of ‘free stuff’ sits on public data, so its defensibility is thin and it’s easy to copy. The lesson isn’t ‘clone Snag’ — it’s ‘treat distribution as a repeatable skill; make apps cheap, fast and disposable, and let the system find the winner.’