01 · DATA ANALYSIS
Finding a $200K creator in a 3,000-follower channel
CONTEXT
AliExpress Korea's influencer group-buying operates on a CPS (Cost Per Sale) model — commissions are paid based on sales performance, with no upfront fees. Actual sales conversion, not channel size, is the core criteria for partner selection.
PROBLEM
Subscriber count indicates reach, but does not correlate directly with purchase conversion. Follower-based channel selection showed low predictive accuracy, and cases of small channels outperforming large ones were repeatedly observed. A quantifiable framework for evaluating sales potential in advance was needed.
MY ROLE
Contribution: ~95%. End-to-end ownership — channel discovery, data tracking, outreach, and contract structure design.
APPROACH
Shifted the channel selection criteria from follower count to the density of purchase-intent audiences. Three indicators were used to evaluate channels.
Inbound Data Tracking
Identified YouTube channels that had reviewed AliExpress-listed products, then tracked actual platform inflow and purchase conversion data through links in video descriptions.
Community Engagement Pattern
Channels that communicated directly with followers through separate channels (e.g., open KakaoTalk chats) tended to attract prospective buyers rather than passive content consumers.
Content Purpose
Channels focused on delivering information and deals — rather than entertainment — showed higher viewer-to-buyer conversion rates.
Applying these criteria, a Tesla accessories channel with approximately 3,000 subscribers was identified. Meaningful sales contribution was pre-confirmed through link tracking data, and the community engagement pattern matched the target profile.
RESULT
Average GMV per campaign: $200K+. Equivalent to the output typically seen from channels with 100K+ subscribers. The same discovery framework applied to a running shoe channel (also several thousand subscribers) produced consistent results of $200K+ per campaign.
LESSON
Follower count is an indicator of reach, not conversion. Purchase behavior is determined by the type of audience a channel attracts — not its size. Viewers who come for information and deals behave differently from those who come for content. This framework was applied to subsequent category expansion (automotive accessories) and produced repeatable results.
The discovery criteria remained personal know-how rather than a formalized scoring system. While insights were shared in team review sessions, a standardized evaluation framework was never documented — which limited scalability across the team.