Analytics forDevelopersand Developing for Analytics
About Me2006-2011: Educated Accounting & Finance2011-Present: Reeducated Marketing & OperationsTwitter: @trent_hauckWork: ...
Two PartsAnalytics(more)Development(less)
Why should you care?
“In God wetrust; allothers mustbring data.”
To do analytics youneed x
Where xis data collection...
Site Analytics Shouldbe a 1st Class Citizenof Development
Collect More ThanYou Need Now
Now Some GA Code<script type="text/javascript">var _gaq = _gaq || [];_gaq.push([_setAccount, UA-31465642-1]);_gaq.push([_s...
Next StepsEvents_gaq.push([_trackEvent, Cat, Act, ‘Label’]);Custom Variables_gaq.push([_setCustomVar, 1, ‘key’, ‘value’, 1])
Where xis data analysis...
Differences in DataSmall Data == Math ProblemBig Data == Engineering Problem
The Math Problem
Descriptive Stats(please compute these)Max, MinQuartilesMeanVarianceMode
Web Stats are EasyA user converts or not... whatare the changes of that?p or q (=1-p)3 users convert or not... what arethe...
Hypothetical WorldsTrials = 100, Size = 100, p = .1
Back to real worldStats: p-bar = .08, SE = 0.027
So then AB Testing500 Trials A Bp 0.1 0.2SE 0.01 0.0195% CI .1 +/- .02 .2 +/- 0.02
The EngineeringProblem
Build Data Pipelines• Repeatable Flows of Data• Handles Initial Analysis For You• Literate Programming
Programming ForData Analysis• Scripting good for Discovery• Larger Jobs need Types• Mapping high dimensional space tolower...
Where xis visualization....
Visualization Types•Distributions•Comparisons•Time Series•Other (Match Domain)
DistributionsSingle Variable: HistogramsMultiple Variables: Scatter plot
ComparisonsCategorical Variables
TimeSeriesX Axis is Time
Match Domain withAnalysis
Where xis storytelling...
Storytelling
3 Temporal Stages1. What happened2. What is happening3. What will happen(Plus a tease)
Start With theSimple StuffFriday Saturday Sunday40º 42º(Why do I livein KC)º
Build Up toComplex Idea
Thanks...Questions?
Upcoming SlideShare
Loading in...5
×

Analytics for Developers - KCDC13

54

Published on

A high level discussion of web analytics.

Published in: Business
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
54
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Analytics for Developers - KCDC13

  1. 1. Analytics forDevelopersand Developing for Analytics
  2. 2. About Me2006-2011: Educated Accounting & Finance2011-Present: Reeducated Marketing & OperationsTwitter: @trent_hauckWork: @AlightAnalyticsOther:Contribute (now and then) to Pandas& StatsModels
  3. 3. Two PartsAnalytics(more)Development(less)
  4. 4. Why should you care?
  5. 5. “In God wetrust; allothers mustbring data.”
  6. 6. To do analytics youneed x
  7. 7. Where xis data collection...
  8. 8. Site Analytics Shouldbe a 1st Class Citizenof Development
  9. 9. Collect More ThanYou Need Now
  10. 10. Now Some GA Code<script type="text/javascript">var _gaq = _gaq || [];_gaq.push([_setAccount, UA-31465642-1]);_gaq.push([_setDomainName, trenthauck.com]);_gaq.push([_setAllowLinker, true]);_gaq.push([_trackPageview]);(function() {var ga = document.createElement(script); ga.type = text/javascript; ga.async = true;ga.src = (https: == document.location.protocol ? https://ssl : http://www) +.google-analytics.com/ga.js;var s = document.getElementsByTagName(script)[0]; s.parentNode.insertBefore(ga, s);})();</script>
  11. 11. Next StepsEvents_gaq.push([_trackEvent, Cat, Act, ‘Label’]);Custom Variables_gaq.push([_setCustomVar, 1, ‘key’, ‘value’, 1])
  12. 12. Where xis data analysis...
  13. 13. Differences in DataSmall Data == Math ProblemBig Data == Engineering Problem
  14. 14. The Math Problem
  15. 15. Descriptive Stats(please compute these)Max, MinQuartilesMeanVarianceMode
  16. 16. Web Stats are EasyA user converts or not... whatare the changes of that?p or q (=1-p)3 users convert or not... what arethe changes of that?p^3 or (p^2)q or p(q^2) or q^3
  17. 17. Hypothetical WorldsTrials = 100, Size = 100, p = .1
  18. 18. Back to real worldStats: p-bar = .08, SE = 0.027
  19. 19. So then AB Testing500 Trials A Bp 0.1 0.2SE 0.01 0.0195% CI .1 +/- .02 .2 +/- 0.02
  20. 20. The EngineeringProblem
  21. 21. Build Data Pipelines• Repeatable Flows of Data• Handles Initial Analysis For You• Literate Programming
  22. 22. Programming ForData Analysis• Scripting good for Discovery• Larger Jobs need Types• Mapping high dimensional space tolower dimensional space... then add
  23. 23. Where xis visualization....
  24. 24. Visualization Types•Distributions•Comparisons•Time Series•Other (Match Domain)
  25. 25. DistributionsSingle Variable: HistogramsMultiple Variables: Scatter plot
  26. 26. ComparisonsCategorical Variables
  27. 27. TimeSeriesX Axis is Time
  28. 28. Match Domain withAnalysis
  29. 29. Where xis storytelling...
  30. 30. Storytelling
  31. 31. 3 Temporal Stages1. What happened2. What is happening3. What will happen(Plus a tease)
  32. 32. Start With theSimple StuffFriday Saturday Sunday40º 42º(Why do I livein KC)º
  33. 33. Build Up toComplex Idea
  34. 34. Thanks...Questions?

×