Resource allocation and revenue optimization for cloud service providers

Online storage and streaming services are surpassing physical media as the predominate means of disseminating and sharing digital contents such as music, documents, photos, and videos. In addition, many software vendors are switching from on-premises installations to web-based rendering for their of...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Decision Support Systems 2015-09, Vol.77, p.55-66
Hauptverfasser: Jhang-Li, Jhih-Hua, Chiang, I. Robert
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Online storage and streaming services are surpassing physical media as the predominate means of disseminating and sharing digital contents such as music, documents, photos, and videos. In addition, many software vendors are switching from on-premises installations to web-based rendering for their offerings. Differentiated pricing, based on tiered service responsiveness and advertisement displays, has been widely adopted by cloud service providers to optimize resource utilization and improve profitability under heterogeneous user demands. We analyze the impact of resource allocation and advertising decisions on provider profit and social welfare when separating premium subscription from more basic offerings. By considering queuing delays and advertising revenues, we suggest conditions under which the service provider should invest in service quality to grow the subscription base, which in turn helps attract more advertisers. We also assess the impact of advertising technology that lessens the users' disutility toward advertisements and increases the likelihood of ads click-through. Finally, we point out when offering free services could be more profitable than charging a subscription fee. •We help content and storage service providers balance their investments in service infrastructure and ads rendering technology.•We show how queueing disciplines can help providers better control service quality.•We optimize service versioning for customers self-select into suitable offerings.•We explain why over-investment in service quality may not attract a desirable customer mix.•We derive conditions under which offering free service or inducements could be preferred over charging subscriptions.
ISSN:0167-9236
1873-5797
DOI:10.1016/j.dss.2015.04.008