Why is a good data scientist a package of professions?
summary
Argues that data scientists must combine technical work with advocacy, design and negotiation, showing concrete fixes: add average active days per day since install to retention reporting, swap distributions for box plots and report bootstrapped percent‑difference, and always include an executive summary with expandable raw plots. Demonstrates a puzzling A/B result where identical groups show opposite TTV because user collection timing differs, explains how continuous collection skews TTV toward day‑0 users versus LTV, and stresses hypothesis, action plan and open communication as operational rules.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 2:33Data scientist as multi-role
- 5:27Data analysis challenges
- 9:17Presenting to stakeholders
- 13:08Designing clear visuals
- 15:14Storytelling with data
- 16:38Soft skills in data science
- 21:18Handling pushback
related talks
32mWhy every developer should be a data scientist
44mCareers in Data Science
58mGrowth & change of careers, organizations & responsibility in data science
54mFrom the Life of a Data Scientist
41mProficiency + Creativity = Happy Data Practitioners
54mThe secret sauce of data science management
43mDomain-driven Data Science
42mHow to be a 10x Data Scientist