Increasing Conversions ThroughBehavioral Targeting            Ross Bauer            Chief Operating Officer            Plu...
Plum Benefits Overview• An employee discount program• Specialize in discounts to ticketed entertainment  (i.e. theater, sp...
Marketing Practices• Email  o   B2B - to program admins  o   B2C - to employees at member firms• Direct Mail• Outbound Cal...
Marketing Practices• B2B Emails  o Target: Program Administrators (i.e. HR    departments)  o List Size: 22,000  o Frequen...
Marketing Practices• B2C Emails (“C” = Employees)  o   Target: Employees at our participating corps  o   List Size: 600,00...
Challenges of B2C (pre-Eloqua)• Content of Email Not Always Relevant to  Consumer• No Visibility of Consumer Behavior on  ...
“Eloqua-izing” Our WebSite
Eloqua’izing Our Web Site• All web users are  written into SFDC  when they register  online• SFDC writes users into  Eloqu...
Increasing ConversionThrough BehavioralTargeting
Behavioral Targeting Overview• Step 1: capture and aggregate user  clickstream (implicit)• Step 2: combine above with what...
Step 1: Capture & Aggregate• Derive implied “area of interest” from  customer click stream:  o Area Of Interest Measured b...
Step 1a: Capture• We use query string parameters to  capture offer views in Eloqua
Step 1a: Great but Now What?
Step 1b: Aggregate• We use query string clusters to group  offers into high level categories like arts &  theater, sports,...
Step 1c: Calculate• We then use profile fields to calculate the  primary area of interest and # of views per group
Step 1d: Move Data from Profile to Contact• We then use a simple program in eloqua  to move area of interest from profile ...
Step 2: Combine• Combine the above implicit data with  other explicit data to form rich user profile         Explicit     ...
Step 3: Eloqua Program• Use Eloqua program to match our offer  emails to the right customer                    Proprietary...
Step 3: The Real Deal
The End Result                 • Recommendations                   are custom and                   based on              ...
Extending the Model
Other Implicit Data We Capture• We use query strings to capture in  Eloqua:  o   Offer Views/Impressions (discussed above)...
Example: Custom Targeting• We just re-signed with the NY Jets and we  want to sell tickets for the 2010 season!• Initial E...
Example: Custom Targeting (cont.)• Those That  o   Viewed, clicked, searched for, and/or      requested       NY Giants  ...
Extending The Visibility ofConsumer Profiles
How we did it: Visibility Offline• Profiles made available in SF via an  integration rule – a little math helps too!
How we did it: Testing• Content Testing: A/B splits• Delivery Testing: Pivotal Veracity/Return  Path
How we did it: Controlling Flow• Content Testing: A/B splits• Delivery Testing: Pivotal Veracity/Return  Path
Results of Implicit vs.Explicit Targeting
Quantifiable Results• Result from April 2009 Campaign
Conclusions1.Behavioral email targeting works2.Does not have to be expensive or time  consuming3.Sharing data to other mem...
Questions?
Upcoming SlideShare
Loading in …5
×

Eloqua's Dirty Little Secret: Behavioral Targeting, A How To Guide

1,480 views
1,379 views

Published on

Published in: Technology
0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
1,480
On SlideShare
0
From Embeds
0
Number of Embeds
16
Actions
Shares
0
Downloads
33
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide

Eloqua's Dirty Little Secret: Behavioral Targeting, A How To Guide

  1. 1. Increasing Conversions ThroughBehavioral Targeting Ross Bauer Chief Operating Officer Plum Benefits
  2. 2. Plum Benefits Overview• An employee discount program• Specialize in discounts to ticketed entertainment (i.e. theater, sports, concerts, family attractions)• Employees from member firms login to browse discount offers• Revenue Model: Fees + Commission on Tix Sales• Primary Goal: Sell Tickets (generate commission)• Marketing team = 5 FTE’s• B2B List (22k), B2C (600k)• Have been using Eloqua for 1-year
  3. 3. Marketing Practices• Email o B2B - to program admins o B2C - to employees at member firms• Direct Mail• Outbound Call Center• Events
  4. 4. Marketing Practices• B2B Emails o Target: Program Administrators (i.e. HR departments) o List Size: 22,000 o Frequency: 3x per month o Content: Offers/Discounts, Service Messages o Goal: “Forward these great offers on to your employees”
  5. 5. Marketing Practices• B2C Emails (“C” = Employees) o Target: Employees at our participating corps o List Size: 600,000 o Frequency  1 regularly scheduled “offer announcement” per month to entire list  Ad hoc targeted messages o Content: Offers/Discounts o Goal: Logon to our site and purchase tickets
  6. 6. Challenges of B2C (pre-Eloqua)• Content of Email Not Always Relevant to Consumer• No Visibility of Consumer Behavior on Web Site for Offline Channels (ie call center)• Long Production Cycles• No Way to Test Emails Before Sending• No Way to Control Flow of Email and Rate of Site Traffic and Inbound Calls
  7. 7. “Eloqua-izing” Our WebSite
  8. 8. Eloqua’izing Our Web Site• All web users are written into SFDC when they register online• SFDC writes users into Eloqua daily• User Login is used as a “form” to capture user behavior in Eloqua
  9. 9. Increasing ConversionThrough BehavioralTargeting
  10. 10. Behavioral Targeting Overview• Step 1: capture and aggregate user clickstream (implicit)• Step 2: combine above with what we already know about user (explicit)• Step 3: Use Eloqua program to match the right offers to the right consumers
  11. 11. Step 1: Capture & Aggregate• Derive implied “area of interest” from customer click stream: o Area Of Interest Measured by number and type of offers viewed on site o Example: If a customer logs onto our site and views 100 offers over a 6-month period:  70 theater offers, 70/100 = 0.70 theater*  20 sports offers, 20/100 = 0.20 sports  10 concerts offers, 10/100 = 0.10 concerts *denotes primary area of interest
  12. 12. Step 1a: Capture• We use query string parameters to capture offer views in Eloqua
  13. 13. Step 1a: Great but Now What?
  14. 14. Step 1b: Aggregate• We use query string clusters to group offers into high level categories like arts & theater, sports, music, family attractions
  15. 15. Step 1c: Calculate• We then use profile fields to calculate the primary area of interest and # of views per group
  16. 16. Step 1d: Move Data from Profile to Contact• We then use a simple program in eloqua to move area of interest from profile record to contact record and then to SFDC
  17. 17. Step 2: Combine• Combine the above implicit data with other explicit data to form rich user profile Explicit Implicit
  18. 18. Step 3: Eloqua Program• Use Eloqua program to match our offer emails to the right customer Proprietary “Magic“ (aka routing algorithm)
  19. 19. Step 3: The Real Deal
  20. 20. The End Result • Recommendations are custom and based on recipients browse pattern
  21. 21. Extending the Model
  22. 22. Other Implicit Data We Capture• We use query strings to capture in Eloqua: o Offer Views/Impressions (discussed above) o Offer Clicks-to-Order o Search Terms (found and not found)• We use cases and web-to-case to capture in Salesforce o Offer Requests
  23. 23. Example: Custom Targeting• We just re-signed with the NY Jets and we want to sell tickets for the 2010 season!• Initial Email Campaign to all those that: o Viewed Jets offer last year (ELQ) o Clicked to buy Jets offer last year (ELQ) o Searched for Jets any time (ELQ) o Requested Jets any time (SDFC)• We also get creative…
  24. 24. Example: Custom Targeting (cont.)• Those That o Viewed, clicked, searched for, and/or requested  NY Giants  Other football teams  Yankees  Mets  Knicks  Nets  All Sports…
  25. 25. Extending The Visibility ofConsumer Profiles
  26. 26. How we did it: Visibility Offline• Profiles made available in SF via an integration rule – a little math helps too!
  27. 27. How we did it: Testing• Content Testing: A/B splits• Delivery Testing: Pivotal Veracity/Return Path
  28. 28. How we did it: Controlling Flow• Content Testing: A/B splits• Delivery Testing: Pivotal Veracity/Return Path
  29. 29. Results of Implicit vs.Explicit Targeting
  30. 30. Quantifiable Results• Result from April 2009 Campaign
  31. 31. Conclusions1.Behavioral email targeting works2.Does not have to be expensive or time consuming3.Sharing data to other members of firm pays dividends by: o Maximizing the impact of the customer hit/conversation o Making you look service-centric
  32. 32. Questions?

×