Big Data, Big Changes:       Data-Driven Product       Development at Etsy                        Strata London, 2012     ...
(The change)Monday, October 1, 12
6 months from inception to launch                        3 engineering teams                        3 product launches    ...
The switch to relevancy           Search, advertising, analytics, and more           Building big data products at Etsy   ...
Monday, October 1, 12
By the Numbers         1.4B pageviews/month         $525MM sales in 2011         40MM unique visitors/month         800k s...
SELECT * FROM listings     WHERE title LIKE     %silver%    ORDER BY creation_tsz    DESC;Monday, October 1, 12
Solving the recency                   problemMonday, October 1, 12
Better buyer                          experience                        Improved search                           relevanc...
Improved advertising                        Search AdsMonday, October 1, 12
More search insight                for sellers                        Shop StatsMonday, October 1, 12
Building big data                            productsMonday, October 1, 12
Build lots of toolsMonday, October 1, 12
Go to the source      Raw data      visit logs      raw query logs      Descriptive statistics      search depth      quer...
Invest in tooling                                      LunrMonday, October 1, 12
Side by side testingMonday, October 1, 12
A/B AnalyzerMonday, October 1, 12
Shop Stats with ReferralsMonday, October 1, 12
Understand tradeoffs                 (most tradeoffs don’t look like this)Monday, October 1, 12
The bike is 20% lighter!Monday, October 1, 12
Queries a, b, & c are better. Queries x & y are not.Monday, October 1, 12
¡Hola!              வணக்கம்!                Guten Tag 你好 !"#$% ‫أ*)م‬                                     ّ               ...
Thinking short-term vs long-term        How will the change affect behavior?        Affect expectations?        User incenti...
Appreciate complexityMonday, October 1, 12
Vs                        Algorithmic ComplexityMonday, October 1, 12
Success / Complexity                              Tradeoff                 4 months                 30 experiments         ...
UI                            June 2005       Complexity                        November 2006                            M...
Communications & Support           ComplexityMonday, October 1, 12
(The change)Monday, October 1, 12
Etsy - Strata NYC                  Continous Experimentation with Continuous                  Deplyoment                  ...
Resources                        Code as Craft blog                        http://codeascraft.etsy.com/                   ...
Monday, October 1, 12
Upcoming SlideShare
Loading in …5
×

Big Data, Big Changes: Data-Driven Product Development at Etsy

1,661 views

Published on

0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
1,661
On SlideShare
0
From Embeds
0
Number of Embeds
99
Actions
Shares
0
Downloads
27
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide

Big Data, Big Changes: Data-Driven Product Development at Etsy

  1. 1. Big Data, Big Changes: Data-Driven Product Development at Etsy Strata London, 2012 Dr. Jason Davis Director of Search & Data jdavis@etsy.com / @jasondavisMonday, October 1, 12
  2. 2. (The change)Monday, October 1, 12
  3. 3. 6 months from inception to launch 3 engineering teams 3 product launches Dozens of experiments Lots and lots of data analysisMonday, October 1, 12
  4. 4. The switch to relevancy Search, advertising, analytics, and more Building big data products at Etsy Invest in tooling Understand tradeoffs Appreciate complexityMonday, October 1, 12
  5. 5. Monday, October 1, 12
  6. 6. By the Numbers 1.4B pageviews/month $525MM sales in 2011 40MM unique visitors/month 800k shops / 150 countriesMonday, October 1, 12
  7. 7. SELECT * FROM listings  WHERE title LIKE  %silver% ORDER BY creation_tsz DESC;Monday, October 1, 12
  8. 8. Solving the recency problemMonday, October 1, 12
  9. 9. Better buyer experience Improved search relevancyMonday, October 1, 12
  10. 10. Improved advertising Search AdsMonday, October 1, 12
  11. 11. More search insight for sellers Shop StatsMonday, October 1, 12
  12. 12. Building big data productsMonday, October 1, 12
  13. 13. Build lots of toolsMonday, October 1, 12
  14. 14. Go to the source Raw data visit logs raw query logs Descriptive statistics search depth query conversion ratesMonday, October 1, 12
  15. 15. Invest in tooling LunrMonday, October 1, 12
  16. 16. Side by side testingMonday, October 1, 12
  17. 17. A/B AnalyzerMonday, October 1, 12
  18. 18. Shop Stats with ReferralsMonday, October 1, 12
  19. 19. Understand tradeoffs (most tradeoffs don’t look like this)Monday, October 1, 12
  20. 20. The bike is 20% lighter!Monday, October 1, 12
  21. 21. Queries a, b, & c are better. Queries x & y are not.Monday, October 1, 12
  22. 22. ¡Hola! வணக்கம்! Guten Tag 你好 !"#$% ‫أ*)م‬ ّ Hello नम#$ 今日はMonday, October 1, 12
  23. 23. Thinking short-term vs long-term How will the change affect behavior? Affect expectations? User incentives? Improved data?Monday, October 1, 12
  24. 24. Appreciate complexityMonday, October 1, 12
  25. 25. Vs Algorithmic ComplexityMonday, October 1, 12
  26. 26. Success / Complexity Tradeoff 4 months 30 experiments 11 winsMonday, October 1, 12
  27. 27. UI June 2005 Complexity November 2006 May 2008 TodayMonday, October 1, 12
  28. 28. Communications & Support ComplexityMonday, October 1, 12
  29. 29. (The change)Monday, October 1, 12
  30. 30. Etsy - Strata NYC Continous Experimentation with Continuous Deplyoment Steve Mardenfeld Scala + Cascading = Scalding Avi Bryant Case Study: What’s a Customer Worth? Roberto Medri October 23-25Monday, October 1, 12
  31. 31. Resources Code as Craft blog http://codeascraft.etsy.com/ Slides (including these) http://www.slideshare.net/etsy Github http://github.com/etsy/ Join in the fun http://www.etsy.com/careers/Monday, October 1, 12
  32. 32. Monday, October 1, 12

×