Your SlideShare is downloading. ×
Everything You Always Wanted to Know About Cohorts (But Were Afraid to Ask)
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×
Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

Everything You Always Wanted to Know About Cohorts (But Were Afraid to Ask)

150
views

Published on

Slides introducing the Wikimedia Foundation's User Metrics API, a platform to extract user activity and engagement metrics and ro run cohort analysis.

Slides introducing the Wikimedia Foundation's User Metrics API, a platform to extract user activity and engagement metrics and ro run cohort analysis.

Published in: Technology

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
150
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
1
Comments
0
Likes
0
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide

Transcript

  • 1. Everything You Always Wanted to Know About Cohorts (But Were Afraid to Ask) Introducing the User Metrics API Ryan Faulkner Dario Taraborelli rfaulkner@wikimedia.org dario@wikimedia.org Monthly Metrics Meeting, March 7 2013
  • 2. ProblemHow do we know if any {program, experiment,new feature} we run to engage new editorsis actually working?
  • 3. +200% increase in key project-level metricdeployment did we really cause this?
  • 4. what if an army of bots didCC BY-NS-SA photo by Andrew Becraft http://www.flickr.com/photos/12426416@N00/232268057
  • 5. what if some external event did http://en.wikipedia.org/wiki/File:Colbert_report.jpg
  • 6. Cohort analysisMeasure the productivity of a cohort: a group of users sharingthe same characteristic or undergoing the same treatmentCC BY photo by Jeremy Mates http://www.flickr.com/photos/jeremymates/2362399109
  • 7. What if we could1. define a cohort from a list of editors Examples • users in an E3 experimental group • students enrolled in a Global Education course • members of a Wikiproject • new editors with 10+ edits in their first month • new users registered on mobile devices • VisualEditor adopters
  • 8. What if we could2. measure the {quality, productivity, retention}of these users via a standard set of metrics threshold measures whether a user reached a given level of activity (a number n of edits) within a specific time t since registration, in a given namespace.
  • 9. What if we could3. compare this cohort to a (meaningful) baseline.
  • 10. Introducing the User Metrics API
  • 11. Introducing the User Metrics API• Standardized set of metrics, backed by extensive research andcrash-tested with multiple experiments run by the E3 team.• Supports cohorts from any Wikimedia project, beyond enwiki(including referral campaigns).• Optimized performance for fast extraction of computationallydemanding metrics from thousands of users.• Returns both individual and aggregate metrics thanks tomultiple aggregator methods. Supports time series.• Easily extensible and fully parametrized (metrics can be fine-tuned, default parameters overridden as needed)
  • 12. Get in touchRyan Faulknerrfaulkner@wikimedia.orgDario Taraborellidario@wikimedia.orgProject homehttps://github.com/wikimedia/user_metrics

×