Stock alerts for a solo reseller of limited-edition products

§01Challenge

One-person business — reselling limited-edition products (mostly sneakers, some collector electronics). Scale: ~40 successful buys per month at $30–200 margin per item. Annual revenue ~$40–70k, run alongside a regular job.

Workflow in 2024: 8 shops (Nike, Adidas, Footshop, Sneakersnstuff, END., Size?, Foot Locker, JD Sports), each with its own "watch" list. Manual F5 every 15–30 minutes. In peak season (drop days) 4–6 hours per day of focused refreshing.

Concrete loss: in one week of March 2024, missed 3 restocks of products worth ~$2,400 in margin combined, because he was at work. Pure missed revenue due to lack of monitoring.

§02Approach

Setup in 1 week. Each of the 8 shops got a dedicated parser (most have JSON-LD or a clear API endpoint). Polling every 60 seconds for "hot" SKUs, 5 minutes for "watch" SKUs. If availability flipped from "out of stock" to "in stock", push notification via Pushover to phone in <30 seconds.

Stack: single Hetzner CPX11 (€4/month), Playwright for 3 shops that require rendering, plain HTTP fetch + JSON parse for the other 5 (open API endpoints). PostgreSQL for state tracking. Bright Data residential for 2 shops with anti-bot.

Critical decisions: NO auto-purchase. Client wanted notification only — buying always manual (on serious sneakerhead forums there is reputational hit for "bot buying", plus every shop's checkout flow is unique and needs human judgment).

§03Outcome

4× more successful buys in the first month post-deployment (16 → 67). Net margin grew from ~$2k/month to ~$5.4k/month.

Significant stress reduction: client admitted "I do not have to refresh during work anymore". Quality of life up, side hustle became sustainable.

Operating cost: €120/month (€4 VPS + €30 proxy + monitoring retainer). ROI measured empirically: first successful buy notification (4 days post-deployment) paid back 3 months of retainer.

System operating 7 months at 99.4% uptime. Three parser updates (each ~1–2h fix), zero critical incidents.

Stack

Metrics

• 8 — Shops monitored
• 40+ — Products watched
• <30s — Notification latency
• 4× — Successful buys lift
• €120 / month — Operating cost
• 4 days — ROI