Where Should You Put Your Data Scientists?

This 2016 Fast Forward Labs presentation discusses the optimal organizational structure for placing data scientists, highlighting three models: stand-alone, embedded, and integrated. Each model has its pros and cons, such as autonomy versus marginalization in the stand-alone model, and utilization versus resentment in the embedded model. The importance of adapting the structure as the company evolves and maintaining focus on the mission is also emphasized.

Transcript

1. Where should you put your data scientists? Daniel Tunkelang
2. Review: why you hire data scientists
3. To improve products
4. To improve decisions
5. Organizational structure matters
6. stand-alone embedded integrated
7. The stand-alone model
8. Pro: gives team autonomy
9. Pro: positions data as first-class
10. Con: risks marginalizing team
11. Con: other teams like self-sufficiency
12. The embedded model
13. Pro: ensures utilization
14. Pro: flexible allocation
15. Con: team resents lack of autonomy
16. Con: management challenges
17. The integrated model
18. Pro: organizational alignment
19. Pro: long-term investment
20. Con: dilution of data science
21. Con: lack of reuse
22. stand-alone: autonomy, data as first class embedded: utilization, flexible allocation integrated: alignment, long-term investment
23. stand-alone: marginalization, isolation embedded: resentment, dual management integrated: dilution, lack of reuse
24. You can mix and match
25. Adapt as company evolves
26. Remember the mission
28. Thank you. Daniel Tunkelang dtunkelang@gmail.com : @dtunkelang