Who Really Drives Sales?

A toy retailer ran a month of TikTok live selling with 10 influencers and needed a fair way to decide who to promote, keep, or dismiss.

The usual trap is to judge on total sales when conditions differ, since some sellers get bigger audiences, better slots, or more popular products. Here the conditions were level. Every influencer ran exactly 120 one-hour sessions across June 2024, so effort was identical. They sold $515,317 between them, and not evenly.

The gap is real. Influencer #3 earned $59,837 and influencer #10 just $38,433, a 1.56 times difference for the same hours, or $499 against $320 per hour live. The question is whether that gap is the person or their circumstances.

The metrics managers usually watch do not explain it. Across the 10 influencers, revenue barely moves with audience size (correlation negative 0.18) or conversion rate (negative 0.22). The only input that points the right way is engagement, and only weakly at 0.38. What revenue actually tracks is units sold, at 0.94: the leaders move more toys, not pricier ones.

Then the fairness checks. Both groups worked identical hour slots, yet revenue still ranged 1.31 and 1.45 times within each group, so the slot cannot be the cause. Selling the same 103 toys, the same people keep the best average rank (#3 ranks 4.85, #10 ranks 6.33), so it is not product luck either.

With effort, slots, and catalogue held equal, the differences trace to the influencer. The recommendation is to promote the four above the $51,532 benchmark, keep the steady middle, and release the two selling 11% and 25% below it, moving 240 monthly hours of prime airtime to proven sellers.