Li et al. 2026 — Endorsement Rate in Influencer Marketing

Summary:

This paper asks whether an influencer’s ad load helps or hurts: does posting lots of sponsored content change engagement on future sponsored posts? It focuses on “endorsement rate” (the share of posts that are ads) and how it predicts later sponsored-post engagement.

Published in the Journal of Marketing by researchers at Hong Kong Polytechnic University and Lingnan University, it’s academic work aimed at explaining the effect and when it changes.

Methods:
• Instagram field study: ~5,000 influencers and 6,869 sponsored posts (one month), using 12 months of prior history to compute endorsement rates.
• Two experiments: 168 Twitter users and 208 Instagram users, manipulating how ad-heavy an influencer seems and measuring engagement intentions/behavior.
• Douyin field study: 147,877 sponsored videos from 1,000 influencers, plus posting histories and comment-text analysis.

Main result: a U-shaped relationship. Engagement is higher when endorsement rates are low or very high, and lowest in the middle.

Why: two opposing forces. More ads can signal brand recognition (boosting engagement), but also trigger suspicion of manipulation (reducing engagement). Suspicion rises early then plateaus; brand-recognition benefits become stronger at high ad loads.

Two factors weaken the U-shape:
• If unpaid posts already cover the same product category, the downside of higher endorsement rates shrinks.
• If the influencer keeps endorsing the same brand, the “brand recognition” upside shrinks.

Caveats: observational platform data isn’t causal despite controls (and fixed effects on Douyin), comment-based mechanisms rely on AI text classification, and Instagram sponsorship measures depend on disclosed ads. Still, the paper is transparent.