Can Genera(ve AI Compete with the Crowd in the First Phase of New Product Development?
Keywords
Loading...
Authors
Issue Date
2025-08-27
Language
en
Document type
Journal Title
Journal ISSN
Volume Title
Publisher
Title
ISSN
Volume
Issue
Startpage
Endpage
DOI
Abstract
This research compares the quality of ideas generated by humans through crowdsourcing with
ideas generated by genera4ve AI (i.e., ChatGPT-4). 200 ice cream flavors (100 crowd-generated and 100
AI-generated) were rated by two independent raters from different ice cream shops on three
dimensions of idea quality: novelty, feasibility and customer benefit. Since the agreement between
raters (IRR) was low, mean idea quality scores were used for the analysis. A MANCOVA tested the
differences between the crowd- and AI-generated ideas, with external mo4va4on (crowd-sample only)
and word count as covariates. The research wanted to test three hypotheses: (H1) crowdsourcing leads
to higher novelty and (H3) customer benefit, while (H2) gen AI produces more feasibile ideas. The
results showed no significant differences in novelty or feasibility between the two groups, but crowdgenerated
ideas did score significantly higher on customer benefit, meaning that only hypothesis 3 was
supported. The findings suggest that ChatGPT can be a fast and efficient tool to generate new and
prac4cal ideas, but the crowd remains important when aligning ideas with real customer needs is most
important
Description
Citation
Supervisor
Faculty
Faculteit der Managementwetenschappen