The cold-start problem in nascent AI strategy: Kickstarting data network effects

Vomberg, Arnd ; Schauerte, Nico ; Krakowski, Sebastian ; Ingram Bogusz, Claire ; Gijsenberg, Maarten ; Bleier, Alexander

[img] PDF
Cold Start_Final.pdf - Published
Restricted to Staff of the repository only

Download (721kB)

URN: urn:nbn:de:bsz:180-madoc-651959
Document Type: Article
Year of publication: 2023
The title of a journal, publication series: Journal of Business Research : JBR
Volume: 168
Issue number: Article 114236
Page range: 1-10
Place of publication: New York, NY
Publishing house: Elsevier
ISSN: 0148-2963
Publication language: English
Institution: Business School > Digitales Marketing u. Marketing Transformation (Vomberg 2021-)
License: CC BY 4.0 Creative Commons Attribution 4.0 International (CC BY 4.0)
Subject: 330 Economics
Keywords (English): artificial intelligence , cold-start problem , data strategy , network effects , value creation
Abstract: While many artificial intelligence (AI) strategies are successful, countless others fail. Why do some strategies succeed while others fail? We adopt a network effects (NEs) perspective to conceptualize AI strategies, highlighting the AI context’s specifics. We argue that nascent AI strategies’ success depends on data NEs: companies establishing a functional “running system” to capitalize on these effects. However, this presents a challenge known as the cold-start problem (CSP), which involves initiating and accelerating a virtuous cycle: more data benefits the AI system, enhancing performance, which then attracts more data. In this paper, we examine the CSP in nascent AI strategy, exploring how it can be understood in terms of its technological and business dimensions and ultimately be overcome to kick-start a virtuous cycle of data NEs. By drawing insights from existing literature and practitioner interviews, we present a research agenda to encourage further investigation into overcoming the CSP.

Dieser Eintrag ist Teil der Universitätsbibliographie.

Das Dokument wird vom Publikationsserver der Universitätsbibliothek Mannheim bereitgestellt.

Metadata export


+ Search Authors in

+ Download Statistics

Downloads per month over past year

View more statistics

You have found an error? Please let us know about your desired correction here: E-Mail

Actions (login required)

Show item Show item