EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

When negative customer engagement turns positive: A four-stage simulation model of engagement evolution in online brand communities

Luning Zang iD, Yuying Liu, Linqiang Wang

PLOS One · 11.11.2026 · 21 (11) · e0360236

DOI: 10.1371/journal.pone.0360236 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, online erschienen am 11.11.2026. Verfasst von Luning Zang und 2 weiteren Autor:innen. Beteiligte Einrichtung: School of Management, Qufu Normal University, Rizhao, People's Republic of China. Hauptthemen laut MEDLINE: Consumer Behavior, Internet. Die Arbeit nennt 20 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Natural Science Foundation of Rizhao, Social Science Planning Project of Shandong Province.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study investigates whether and when service recovery can convert negative customer engagement (NCE) into positive customer engagement (PCE) in online brand communities, and how early value conditions shape later engagement trajectories. Using Xiaomi Community as the focal context, we develop a four-stage threshold-based simulation framework in which consumer utility is treated as a latent evaluation antecedent rather than customer engagement itself. An explicit non-engagement (NE) interval is positioned between the negative and positive thresholds so that moderately evaluated users are not forced into PCE or NCE. Price, product quality, service quality, and personal needs met are retained as the focal antecedents. Core scenarios are repeated 1,000 times with recorded random seeds and 95% simulation uncertainty intervals, and robustness is examined across conversion rates, thresholds, initial conditions, coefficient distributions, sign-allocation rules, and simulation horizons. Price and product quality show strong and opposite effects on the PCE-NCE balance, and a direct price-by-quality analysis indicates that higher quality attenuates the adverse effect of higher price. Service-quality effects remain under both Stage-4-only and continuous-service architectures, but the results do not support a general claim that later-stage service investment is superior to earlier service. Targeted recovery applied to NCE users produces a stronger and more stable NCE-to-PCE repair effect. The results clarify the assumptions, robustness, and managerial boundary conditions of the stage-based model. Because coefficient sign allocation remains a major structural uncertainty, the reported quantities are interpreted as scenario outputs and robust directional comparisons rather than forecasts of actual PCE, NCE, or NE population shares.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Consumer BehaviorInternetHumansComputer SimulationCommerce

Autor:innen und Institutionen

  • School of Management, Qufu Normal University, Rizhao, People's Republic of China.
  • School of Management, Shandong University, Jinan, People's Republic of China.
  • School of Tourism, Shandong Women's University, Jinan, People's Republic of China.

ORCID-Profile: Luning Zang

Förderung

  • Natural Science Foundation of Rizhao
  • Social Science Planning Project of Shandong Province

Zitieren

APA

Zang, L., Liu, Y., & Wang, L. (2026). When negative customer engagement turns positive: A four-stage simulation model of engagement evolution in online brand communities. PLOS One, 21(11), e0360236. https://doi.org/10.1371/journal.pone.0360236

Vancouver

Zang L, Liu Y, Wang L. When negative customer engagement turns positive: A four-stage simulation model of engagement evolution in online brand communities. PLOS One. 2026;21(11):e0360236. doi:10.1371/journal.pone.0360236

BibTeX
@article{Zang2026,
  author = {Zang, Luning and Liu, Yuying and Wang, Linqiang},
  title = {When negative customer engagement turns positive: A four-stage simulation model of engagement evolution in online brand communities},
  journal = {PLOS One},
  year = {2026},
  volume = {21},
  number = {11},
  pages = {e0360236},
  doi = {10.1371/journal.pone.0360236}
}
Kostenlos starten →