Augmenting Human Decision Making with Data Science
summary
Data and concrete results: a styling app for Stitch Fix matches 3,000 stylists to customers, filters inventory, computes match scores, and shows ranked, smaller lists (an A/B test finds a 50‑item sweet spot); 95% of decisions rely on fast intuition yet algorithms train and guide stylists, reducing time and errors, raising keep rates and lowering labor cost. A tiny fraction of fixes get reassigned when the algorithm overrides a low‑score stylist selection; dashboards and feedback loops continuously improve models.
AI-written summary · about
chapters (14)
- 0:00Start
- 4:01What are the limitations of human decision making?
- 4:37Intuition & feelings
- 4:56Interpretation of our surroundings
- 5:4495% of human decisions are made in System
- 10:48INCONSISTENT BIASED INFO OVERLOAD
- 12:51STITCH FIX Case Study
- 14:01How can human decisions be augmented by data science?
- 19:53Stylists use feedback to train to make more accurate decisions in the future
- 24:25FOLLOW YOUR INTUITION
- 25:35What is the future of humans and data science?
- 26:21Predictive & intuitive
- 27:01What is the business impact of system 3?
- 28:35Humans lack the ability to process large volumes of information.
related talks
42mAugmenting Human Decision Making with Data Science
20mBrad Klingenberg, Director of Styling Algorithms, Stitch Fix
25mPersonalization and Retail: Lessons from Stitch Fix
02mDBTB IN Eric Colson r
32mScaling Personalization: DevOps at Stitch Fix
18mSynthesizing human and machine capabilities
42mAn Algorithm of Style
39mDecision Making Based on Machine Learning at Outfittery