Half Stack Data Science S04E04 - Better survey design with AI - with Kirsten Hill
Jun 22, 2025
Kirsten Hill, a researcher and founder of the innovative survey platform bisque, discusses the critical shortcomings of traditional survey methods. She highlights how many surveys fail to yield actionable insights due to poor design. Kirsten shares her mission to revolutionize research with AI-driven tools that simplify survey creation. She also provides valuable advice for non-programmers looking to leverage AI in their projects, emphasizing accessibility and best practices in data collection. Tune in for her fresh perspective on effective survey design!
AI Snips
Chapters
Books
Transcript
Episode notes
Quitting Government Job Early
- Kirsten quitted a government job after four months due to lack of motivation and too many holidays delaying the work.
- She realized she preferred entrepreneurial and tech startup environment despite family concerns about job security.
Surveys Often Yield Garbage Data
- Many surveys are poorly designed with double-barreled or unclear questions, yielding unusable data.
- Organizations often spend money on such surveys without meaningful plans to act on the results.
Survey Data Driven by Boss Pressure
- Many organizations gather survey data because of pressure from superiors rather than effective research goals.
- This leads to poorly thought-out surveys whose results have limited practical use.
