Taking on freelance work, in public
Dated notes from starting out on Taiwan freelance platforms (Tasker first). No client names, case details or deal prices. Income so far: NT$0 (real data — no paid client yet).
A pricing rule before sending any proposal
What was done
- Set a quoting rule: cost (working time + a 20% revision buffer + AI/tool fees + points cost) × 1.2. Each proposal goes out only with its cost sheet approved.
- Sent 9 Tasker proposals that day (5 at noon, 4 in the evening); the platform showed each as successfully submitted.
- Also drafted 5 service listings, not published yet.
What was learned
- With zero reviews, peers lower the client's risk with a small trial, a 50% deposit and a fixed number of revisions — not by cutting price.
- Quoting too low on a large job can read as unprofessional; low pricing should depend on job size.
Numbers (real data)
- Markup × 1.2 (cost + 20%) — own pricing rule.
- 9 proposals sent (5 noon, 4 evening); 0 deals so far, income NT$0 — own records.
- 5 service listing drafts — own daily log.
Bidding points are locked, not spent
What was done
- Read Tasker's official help and pricing pages on proposal points and wrote a bidding rule from them.
What was learned
- Points put on a proposal are refunded 100%, won or not — returned when the client accepts, declines, closes the job, or leaves it unread/unanswered for 2 days. So a bid locks points for about 2 days rather than spending them.
- The real cost is opportunity cost (points locked away from other bids) and any credit that expires unused. Refunded credit keeps its original expiry date.
- In the priority-proposal area, ranking is by points put in only; ties go to whoever proposed first.
Numbers (real data)
- Refund rate 100%; refund after 2 days unread — Tasker official help pages.
- Bid per proposal: min 100, max 9,999 points; official suggestion about 200 per job — Tasker official pages.