Live streaming channel strategies in the presence of the dual effect of influencers

Live streaming (LS) has become a vital sales channel for brand owners. This study evaluates three LS strategies: self LS (S model), where the brand sells directly through its own account; influencer LS (I model), where products are sold via an influencer's account; and hybrid LS (C model), which combines both approaches. Our findings reveal that the choice of strategy is strongly influenced by the influencer's fan effect. A high fan effect favors the I model, followed by the C model, while a low fan effect favors the S model. As customer viewing costs increase (e.g., due to shorter streaming durations), sales prices rise in the C model but drop in the I model. However, the brand owner's profit increases in both I and C models. We also find that moderate reputation spillover benefits both the brand owner and the influencer, but the influencer may not always deliver maximum effort in the I model. Additionally, consumer welfare is always highest when the C channel is used, and consumer interests may not always align with those of the brand owner and influencer. In the extensions, we consider platform streaming algorithms and external competition, further enriching the research conclusions. This study offers guidance to brand owners in selecting LS strategies, highlighting the trade-offs between self-live streaming and influencer collaboration.