Loading
Loading
Taste
Since 2013, 491 collectors contributed 7,240 weekly ‘taste’ collections of designs they found desirable, each coded against 52 dimensions of desirability — then clustered and ranked by distinctiveness. The DL was founded in 2013 by Professor Altringer Eagle at Harvard. She continues to run it since 2025 at Dartmouth.
226
collectors profiled
4
taste clusters
52
dimensions
8%
silhouette
The Observer
Form, elegance, media
The Pragmatist
Efficiency, clever solutions, tech
The Idealist
Social impact, belonging
The Tastemaker
Fashion, brand strategy
Each row is one cohort year, showing the domain mix of what collectors chose.
The first cohort. Students finding their voice in a brand-new assignment format, still figuring out what it means to collect desirable things.
The assignment settles into its rhythm. Students begin to notice patterns in their own taste, and the collections start to feel personal.
Social media products surge in popularity. Instagram, Snapchat, and Vine reshape what students think of as desirable design.
A broader aesthetic palette emerges. Architecture, food, and fashion compete with tech for attention. Collections become more diverse.
The tech obsession peaks. Smart home devices, wearable tech, and app-first products dominate — this cohort sees the future as designable.






A pivot toward sustainability and social impact. Students increasingly choose products that signal values, not just function.



The collections mature. Students show more design literacy, referencing materials, craft, and manufacturing in their reviews.
Pandemic year. Collections shift toward home, wellness, and digital experience. The domestic sphere becomes the stage for desirability.






The final cohort in this dataset. A synthesis of trends — tech, sustainability, wellness, and craft all coexist.
The final cohort in this dataset. A synthesis of trends — tech, sustainability, wellness, and craft all coexist.
Each dot is a collector, positioned by how far their taste vector deviates from the population mean (cosine distance percentile).