Three years ago I watched a seller in Hangzhou move $40,000 of skincare in a four-hour stream from a room the size of a parking space. No agency, no ad budget - just a phone, a ring light, and a script she had clearly run a thousand times.
That script is the product. Not the product she was selling - the script itself, the pacing, the offer structure, the way she re-opened every 90 seconds for the people who had just arrived. In China, that operational knowledge is an industry. In the US, where TikTok Shop is growing fastest, most sellers are still improvising it alone.
So I spent the last weeks building Live Selling Playbook - an open knowledge base that translates China's live-selling operating system into English, for US TikTok Shop sellers.
What actually transfers (and what doesn't)
The surprising finding from researching this: almost everything transfers, but nothing transfers automatically.
The mechanics are identical - TikTok's live algorithm worldwide rewards the same signals: early watch time, session length, comment density, returning viewers. What changes is the wrapper:
- Offer cadence. Chinese rooms run a 5-minute loop: hook, demo, offer, proof, reset. US sellers tend to narrate ("welcome back everyone!") instead of re-selling. The loop isn't a culture thing - viewers arrive continuously everywhere, and every arrival needs the offer within their first minute.
- The trust stack. Chinese buyers ask logistics questions (real? shipped? refund?); US buyers add compliance anxiety (is this an ad? is that claim legal?). Same playbook, extra checklist layer.
- Follower myths. The most-shared "requirement" articles are wrong in a useful way: sellers on TikTok Shop can go live with zero followers through the shop channel, while creators need 1,000. I wrote up the full rule set with sources and dates here.
The part nobody wants to hear: income math
Every "how much do live sellers make" thread attracts two answers - screenshot flexes and doom. The arithmetic is more interesting and more useful:
income per session = viewers x conversion rate x avg order value x margin
A 50-viewer room converting at 4% outsells a 300-viewer room at 0.5% - and the smaller room is easier to run because you can answer every comment by name. The leverage isn't viewer count, it's conversion discipline. I put realistic 2026 monthly ranges and the full formula in this guide because I'd rather under-promise than join the hype economy.
The build (for the devs here)
Since this is DEV: the site is Next.js on Netlify, content lives in typed TypeScript files (one interface, 80+ entries), Supabase handles accounts and PayPal webhook-driven memberships, and the whole SEO layer is deliberately boring:
- one page = one question keyword, question keywords from real US search phrasing
- every claim carries a "last verified" date - the differentiator in a niche full of undated advice
- free guides stay free forever; membership unlocks the operator layer (scripts, benchmarks, playbooks)
The content pipeline is the fun part: a standing batch process that translates and re-verifies China's live-commerce knowledge into dated, sourced English guides. Turns out a knowledge base is one of the few businesses where "boring and correct" is a moat.
If you're selling on TikTok live
Start with the free layer - the full library of guides covers setup, timing, moderators, troubleshooting dead rooms, and the income math. If you've run lives yourself: what's the thing you wish someone had told you before your first session? Genuinely asking - the best entries in the knowledge base come from operator war stories.
