Google Analytics Attribution
2 likes1,084 views
This document discusses Google's multi-channel attribution tool. It notes that marketing teams need a holistic view of how customers engage across different channels over time. Google's tool uses advanced statistical modeling to attribute conversions to the appropriate channels based on each customer's ad exposure history. This allows teams to better understand and optimize the performance of their multi-channel campaigns.
Google Analytics Attribution
- 1. CHANNEL ATTRIBUTION WITH GOOGLE Thomas Bosilevac Director of Analytics @bosilytics Thursday, May 17, 2012
- 2. We are not seats or eyeballs or end users or consumers Thursday, May 17, 2012
- 3. nor are we impressions or clicks or conversions Thursday, May 17, 2012
- 4. WE ARE HUMAN BEINGS AND OUR REACH EXTENDS YOUR GRASP. Thursday, May 17, 2012
- 5. DEAL WITH IT -Cluetrain Manifesto 1999 Thursday, May 17, 2012
- 6. WELCOME MULTI-CHANNEL ATTRIBUTION Thursday, May 17, 2012
- 7. We Live in an Increasingly Complex, Multi-Channel Ecosystem Thursday, May 17, 2012
- 8. Optimizing Multi-Channel Marketing Programs can be a Major Challenge Teams on islands with isolated data often achieve sub-optimal results -*./*.%0"& '*+$%,& 1.%2#& '()& 5-*6& 34& !"#$%& '(!& Thursday, May 17, 2012
- 9. Teams Need a Holistic View of How their Customers Engage with Media Campaigns We need to understand all of the campaign touch points in the customer lifecycle and how much they contribute to helping us reach our goals Thursday, May 17, 2012
- 10. Siloed Teams and Data Forces Use of “Last Ad” Standard But campaigns are made up of overlapping ads that reach consumers multiple t across multiple channels, over extended periods of time Banner Rich Media & Banner Banner Search Yahoo Sponsorship MSN CNet Sky Sports Google $ $ $ $ $ $ Thursday, May 17, 2012
- 11. Without a Holistic View of Where Customers Engage, Teams Enter a Vicious Cycle Optimizing media without understanding what’s really driving towards business goals will degrade efficiency and performance Thursday, May 17, 2012
- 12. Evaluating Every Touch Point Helps Teams Have a More Holistic View of the Customer Using advanced attribution modeling, Yahoo drove 15% more conversions than it did with the last ad model Thursday, May 17, 2012
- 13. SO WHAT CAN I DO TODAY? Google Multi-channel conversions Thursday, May 17, 2012
- 14. First Interaction Banner Yahoo Rich Media & Sponsorship MSN Banner CNet Banner Sky Sports Search Google Conversion $ Thursday, May 17, 2012
- 15. WHAT IS A CONVERSION? ANY Transaction or Goal 15 Thursday, May 17, 2012
- 16. SO REALLY, HOW ARE MY BANNERS? Thursday, May 17, 2012
- 17. SO REALLY, HOW ARE MY BANNERS? 17 Thursday, May 17, 2012
- 18. SO REALLY, HOW ARE MY BANNERS? Thursday, May 17, 2012
- 19. Thursday, May 17, 2012
- 20. SO REALLY, HOW ARE MY BANNERS? Thursday, May 17, 2012
- 21. OTHER REPORTS Time Lag Path Length Days to convert Interactions to conversion Thursday, May 17, 2012
- 22. CONVERSION SEGMENTS Thursday, May 17, 2012
- 23. CHANNEL DESCRIPTIONS display, cpm, AdWords “content” through ad Display distribution Paid Search AdWords and/or “cpc”, “ppc” Organic unpaid search engines or “organic” Social ~400 social networks (no tag) Referral any other site not listed Email “email” Direct Through URL, Bookmark or “(not set)” Other Anything else (custom medium / source) Thursday, May 17, 2012
- 24. SO WHAT CAN I DO TODAY? Google Multi-channel conversions Thursday, May 17, 2012
- 25. Advanced Attribution Allows Teams to Weight KPI Contributions to all Ads Seen For the campaign lifetime, many ads will have a role in driving conversion 21 14 7 1 J.G+$ $ 6(7 6(7 6(7 6(7 6(7 8$ !$ 9$ :$ ;";!$ $ 3(45$%& =;;#$ ;#$ ;#$ ;#$ ;#$ $ %,-./012)$ %&'()*+& %,-./012)$ ;";8<$ !"!#$ ;"=<<$ !"!#$ ;";>!$ !"!#$ ;"=?9$ ;"?!?$ !"!#$ $ %)*A2- J-.K$%&'.42- @(A22 %C3 E.+F$ L+-*A()5$M25+N BCD B+G(-H+1)I E.+F$ E.+F$ E.+F$ Thursday, May 17, 2012
- 26. A Customized Statistical Model will Attribute Each Conversion to the Appropriate Channel Probability to convert as a function of ad exposure ln ( = b + m1x1 + m2x2 + m3x3 + … + mixi • The ad history is defined by variables xi which define ad characteristics (e.g., display vs. search) or behavior (e.g., click vs view). • The coefficients (mi) define the weightings associated with each ad characteristic/behavior • The model will incorporate the coefficients for each ad to determine 26 how much weight to apply to each ad in a converter’s exposure history Thursday, May 17, 2012
- 27. A 360 Degree Marketing Intelligence Platform can Help with Holistic Customer Insights The platform integrate data from online, customer data, third parties, and offline media channels 27 Thursday, May 17, 2012
- 28. About Digitaria’s Analytics Department Analytics Solutions Select Vendor Partnerships • Measurement Strategy • Analysis • Testing • Optimization • Decision Support • Dashboards Our Global Analytics Clients Thursday, May 17, 2012
- 29. Q&A Thursday, May 17, 2012