Harnessing AI to Enhance Crowdsourcing during Ideation
In a groundbreaking discovery, researchers have learned to harness the power of artificial intelligence (AI) to enhance the crowdsourcing process during ideation. By developing a simple model, they can now focus on high-quality ideas while filtering out the less promising ones.
Managing the Model for Optimal Balance
Managers can modify this model to determine an optimal balance between disregarding poor ideas and retaining good ones. Crucially, the study also identified a new predictor that sifts out unconventional ideas and preserves more diverse and rich suggestions.
Paper Insights and Authors
A recently published paper in the journal Marketing Science, titled “Can AI Help in Ideation? A Theory-based Model for Idea Screening in Crowdsourcing Contests,” explores these findings. The paper’s authors are J. Jason Bell from the University of Oxford, Christian Pescher from Universidad de los Andes in Chile, Gerard Tellis from the University of Southern California, and Johann Füller from the University of Innsbruck in Austria.
Challenges in Crowdsourcing
Crowdsourcing contests can generate hundreds or thousands of ideas for businesses seeking innovative products or services. Sifting through these ideas manually is time-consuming, inconsistent, and prone to human biases. The team focused on AI as a potential solution, given its capability to improve the process’s efficiency.
AI as a Solution
As stated by Gerard Tellis, “Idea generation and screening play a pivotal role in marketing success because they initiate new product development. They form an essential part of the ‘fuzzy front end’ of new product development, which serves as a critical leverage point.”
AI Model Training and Data
The researchers used data from Hyve, an innovation company specializing in crowdsourcing, to train their AI model. By setting the accuracy threshold desired by Hyve’s clients, they demonstrated the AI model’s potential using data containing 4,191 ideas from 21 crowdsourcing contests.
Findings and Advantages of AI
- AI models have relatively low operating costs once established
- They are less likely to be subjected to human biases
- Offer privacy, intellectual property protection, and aren’t affected by human exhaustion
- Provide transparency in evaluations
Conclusion and Potential Impact on Crowdsourcing
Though human involvement will still be necessary, particularly in the selection phase, AI systems can swiftly screen and narrow down ideas during crowdfunding. If leveraged properly, this technology might eventually eliminate the reliance on human idea generators, rendering the crowdsourcing process itself obsolete.
