Consumer classifieds marketplace
Took a web-only marketplace to web, iOS and Android, added a second payment rail and one experiment framework, and grew collected revenue from $2.42M a year to $6.33M.
Adly portfolio company, 2023 to Aug 2026. I own the product. Updated .
Context
A consumer classifieds marketplace in Adly's portfolio that earns from subscriptions and advertising. I own the product, the developers who build it, and its go-to-market with the growth, moderation, ad and support teams.
When I took it over it was one web product on one payment rail, with a codebase that was hard to change safely and no shared way to test pricing.
The problem
Revenue depended on a single web funnel and a single billing provider. There was no app for mobile users. The moderation and ad teams needed tools faster than they were getting them, and pricing changes had no controlled way to be tested.
What I built
- Web, iOS and Android storefronts
- Two payment rails, rolled out by percentage
- One cohort framework for every monetization test
- Promote to core, or kill inside the window
Winners become the baseline for the next test.
Mobile
Launched on Google Play on May 20, 2025 and the App Store on Jun 3, 2025, starting with about 100 users per build. Since then, a release roughly every month, with Android on 25%, 50%, then 100% phased rollouts.
A second payment rail, no risky cutover
Routed 5% of upgrades, then 10%, then about 50 to 60% at the upgrade click, then lapsed users from the original rail. By Aug 2026 it carried 32.5% of web cash.
One experiment framework
10 tests reached a decision. Four were killed inside their windows, including a 100% price expansion reverted within 7 days. Renewal price tests raised revenue per assigned user in all four cohorts: +65%, +100%, +29% and +54% in an early read.
AI trust and safety
AI image and text moderation, a four-level shadowban system and location-to-IP checks. AI reviewed all 27,348 published stories and confirmed 4,648 violations before the feature opened up. A backtested chat cascade covers 100% of messages for an estimated $570 a month, against $1,825 for today's sampled subset.
Tools for the ad team
About 15 new tools and ad units. One banner alone created 6 new sellable ad units.
A faster, safer delivery loop
About 66% more tickets shipped per month (826 over 19 months against 471 over the prior 18), with rollbacks flat at 1.7%. An automated PR review bot has logged 2,508 findings across 258 PRs.
Results
Collected revenue by year
| Year | Relative size | Collected revenue | Growth |
|---|---|---|---|
| 2023, web only | $2.42M | ||
| 2024, web only | $4.02M | +66% | |
| 2025, web and apps | $5.43M | +35% | |
| Last 12 months, to Aug 2026 | $6.33M | +28% | |
| 2023 to last 12 months | +162% |
Month by month, since Mar 2025
| Measure, monthly | Mar 2025 | Aug 2026 | Change |
|---|---|---|---|
| Collected revenue, all rails | $410.4K | $560.9K | +36.7% |
| Contracted MRR | $263K | $536K | +104% |
| Recurring web subscribers | 47,458 | 77,966 | +64% |
| MRR per recurring web subscriber | $5.53 | $5.97 | +8% |
| Mobile app revenue, gross | $0 | $87.7K | New line |
- Collected revenue is cash that cleared, net of refunds. 2023 and 2024 are web only because the apps did not exist yet.
- 2023 includes the early ramp of the billing system, so growth measured from 2023 reads higher than the steady-state rate.
- Part of the mobile step-up in Mar 2026 shifted from web billing rather than being new revenue.
- Contracted MRR normalizes yearly plans to monthly and excludes one-time Lifetime purchases.
- View-to-purchase conversion rose from 1.90% to 5.57% in Dec 2025 and held for 8 months. The driver isn't isolated, so I don't claim it.