The Circulate and Recapture Dynamic of Fan Mobility in Agency-Affiliated VTuber Networks

TL;DR

The paper addresses how VTuber agency affiliation (MCNs) shapes fan attention dynamics by linking micro-level fan trajectories (retention, oshi switching, inactivity) to meso-level audience-overlap networks. Using a three-year, fan-centered panel across 458 channels and monthly overlap graphs, the authors demonstrate a 'circulate and recapture' dynamic within agency portfolios: fans stay engaged and frequently shift among affiliated creators, sustaining participation without locking into a single channel. Network analyses reveal that affiliated segments develop denser local neighborhoods and more cohesive cores, even as global edge density converges with independents, suggesting portfolio-level coordination that channels attention. The contribution is a reusable measurement framework that connects micro behaviors to meso network structure, offering insights for creator labor, influencer marketing, and platform governance, while noting that the results are descriptive and not causal.

Abstract

VTuber agencies -- multichannel networks (MCNs) that bundle Virtual YouTubers (VTubers) on YouTube -- curate portfolios of channels and coordinate programming, cross appearances, and branding in the live-streaming VTuber ecosystem. It remains unclear whether affiliation binds fans to a single channel or instead encourages movement within a portfolio that buffers exit, and how these micro level dynamics relate to meso level audience overlap. This study examines how affiliation shapes short horizon viewer trajectories and the organization of audience overlap networks by contrasting agency affiliated and independent VTubers. Using a large, multiyear, fan centered panel of VTuber live stream engagement on YouTube, we construct monthly audience overlap between creators with a similarity measure that is robust to audience size asymmetries. At the micro level, we track retention, changes in the primary creator watched (oshi), and inactivity; at the meso level, we compare structural properties of affiliation specific subgraphs and visualize viewer state transitions. The analysis identifies a pattern of loose mobility: fans tend to remain active while reallocating attention within the same affiliation type, with limited leakage across affiliation type. Network results indicate convergence in global overlap while local neighborhoods within affiliated subgraphs remain persistently denser. Flow diagrams reveal circulate and recapture dynamics that stabilize participation without relying on single channel lock in. We contribute a reusable measurement framework for VTuber live streaming that links micro level trajectories to meso level organization and informs research on creator labor, influencer marketing, and platform governance on video platforms. We do not claim causal effects; the observed regularities are consistent with proximity engineered by VTuber agencies and coordinated recapture.

Figures (5)

• Figure 1: Commitment metrics by affiliation (Agency vs. Independent). Each panel shows a violin plot. Metrics: (a) two month retention; (b) three month retention; (c) longest active run; (d) oshi switch probability.
• Figure 2: Sankey diagrams of fan flows over a three month horizon for cohorts from (a) agency and (b) independent origins. At each monthly time point (T1 to T3), fans are classified into one of four states: Same, Retain, Cross, or Drop.
• Figure 3: Time series of network metrics for the Agency and Independent segments. (a) Network density and (b) average degree over time.
• Figure 4: Representative snapshots (January of each year) for Independent and Agency segments. Force directed layouts with fixed seeds; node size $\propto$ degree, edge width $\propto S_{ij,m}$. Isolated nodes are omitted for readability.
• Figure 5: Node-level metrics on the unified audience--overlap network, by affiliation (Agency vs. Independent). Each panel shows a violin plot with creator-level points and a horizontal line indicating the mean. Metrics: (a) weighted degree (sum of Simpson weights); (b) degree; (c) local clustering coefficient; (d) unweighted betweenness centrality.