How to attribute multi-channel donations
•
1 j'aime•1,013 vues
Multi-channel campaigns generally mean better results for your organization. But when a single constituent receives consistent messaging through multiple channels, how does your organization know which source to attribute the response? We’ll discuss our attribution model for integrated multi-channel campaigns that helps organizations determine which channel gets “credit” for which conversions. This powerful insight allows organizations to more accurately allocate budget to the channels that their constituents are most responsive to and understand how one channel influences another.
Recommandé
Recommandé
Contenu connexe
Tendances
Tendances (8)
Similaire à How to attribute multi-channel donations
Similaire à How to attribute multi-channel donations (20)
Dernier
Dernier (20)
How to attribute multi-channel donations
- 1. Where did that gift come from? Attributing multi-channel gifts #attribution Angela Struebing The Lukens Company Amy Katherine Radick Art Institute of Chicago
- 2. Philadelphia Museum of Art / February 14, 2013 attribution 101
- 3. Where did that gift come from? #attribution • What is attribution? • Various methods • “In theory” • Art Institute of Chicago Case Study • Best Practices for Multi-channel campaigns 3 web ads your website teleseminars/ webinars direct mail email blogs broadcast media pr print ads social media sites videos mobile telemarketing outdooradvertising Where did that gift come from? Attributing multi-channel gifts
- 4. Where did that gift come from? #attribution According to Webster… 1.to regard as resulting from a specified cause; consider as caused by something indicated (usually followed by to): She attributed his bad temper to ill health. 2.to consider as a quality or characteristic of the person, thing, group, etc., indicated: He attributed intelligence to his colleagues. 3.to consider as made by the one indicated, especially with strong evidence but in the absence of conclusive proof: to attribute a painting to an artist. 4.to regard as produced by or originating in the time, period, place, etc., indicated; credit; assign: to attribute a work to a particular period 4 at·tri·but·tion [v. a-truh-byoot-shun]
- 5. Where did that gift come from? #attribution Why does it matter? • Prospects don’t see a single message • There are multiple channels to devote resources to • Each channel has different metrics ($, emails, sign-ups, etc.) • Lifetime Value of Donors • Channel decay 5
- 6. Where did that gift come from? #attribution 60 days 30 days 2 weeks 1 week mail digital ads phone email Response Decay
- 7. Where did that gift come from? #attribution Choose your own adventure email mail telemarketing ads email mail telemarketing email ads 7
- 8. Where did that gift come from? #attribution Neither factor in organic joins Neither gives credit to marketing touches (no ROI tied to ads) Neither allows for fractional allocation for multiple touches Current Practices Confirmed Response – using mail/promo codes to assign channel (i.e. mail, phone, lobby, email, online) • Only get results when reply device is used, email links are clicked or ad landing pages are completed • Lots of unidentified responses Matchback – when no code is present, we match respondent data with outgoing messages • Fewer unknown respondents • Follow-ups are most likely given undue credit 8
- 9. Where did that gift come from? #attribution • First Touch/Last Touch – whatever the first or last channel of outward communication was with the respondent receives 100% credit for donation • Linear – each touchpoint in the conversion path shares credit for the donation • Tiered– the touchpoints closest in time to the donation get most of the credit • Position Based – more weight is given to the first and last touchpoints with midstream touchpoints marginalized 9 Fractional Allocation
- 10. Where did that gift come from? #attribution 10 Mail Email Phone Email 2 Cookie Ad First Touch $100.00 – – – – Fractional Allocation $17.00 $18.60 $20.00 $20.90 $23.00 Weighted Allocation $15.64 $13.59 $39.85 $15.17 $15.75 One big math problem
- 11. Where did that gift come from? #attribution • Drop date*, not in home • Doesn't include multis • Could have received multiple display impressions (7-8 recommended) • Doesn’t account for varying tail lengths • Still doesn’t factor in organic donations 11 Mail Email Phone Email 2 Cookie Ad Donation $100 3/18/13* 3/27/13 4/1/13 4/3/13 3/18-4/12/13 4/10/13 Frequency 1 1 1 1 1 Recency 0.5 0.6 0.7 0.8 1 Weighting 1.5 1.6 1.7 1.8 2 8.6 Attribution 17% 18.6% 20% 20.9% 23% Allocation $17 $18.6 $20 $20.9 $23 $100 based on 1 month shortened timeframebased on 1 month shortened timeframebased on 1 month shortened timeframebased on 1 month shortened timeframebased on 1 month shortened timeframebased on 1 month shortened timeframebased on 1 month shortened timeframe Taking it a step further with weights
- 12. Where did that gift come from? #attribution There is a better way 12 membership purchase control mail email cookies facebook
- 13. Philadelphia Museum of Art / February 14, 2013 multi-channel attribution
- 14. Where did that gift come from? #attribution 14 Multi-channel AttributionControl Groups • No Touch – randomly selected targets withheld from all direct touches • Single Touch – randomly selected targets given a single touchpoint (i.e. mail, email, Facebook Custom Audiences, Cookie Matches, phone) Bucket Test • Multi Touch – combination touchpoints (i.e. mail and email) to test if effects are additive versus multiplicative compared to the control *For best results, no other correspondence should be directed to these audiences for 60 days after the mailing!
- 15. Where did that gift come from? #attribution 15 Channel Weights Mail Email Phone Cookie Ad Single Touch Response Rates 1.25% 0.40% 8% 0.10% Control (organic) 0.09% 0.09% 0.09% 0.09% Difference* 1.16% 0.31% 7.91% 0.01% Weight 0.49 0.13 3.37 0.004 *collective difference 2.35%
- 16. Where did that gift come from? #attribution 16 Mail Email Phone Email 2 Cookie Ad Donation $100 3/18/13* 3/27/2013 4/1/2013 4/3/2013 3/18-4/12/13 4/10/2013 Frequency 1 1 1 1 1 Recency 0.5 0.6 0.7 0.8 1 Attribution Weight 0.49 0.13 3.37 0.13 0.004 Adjusted Weighting 1.99 1.73 5.07 1.93 2.004 12.724 Adjusted Attribution 15.64% 13.60% 39.85% 15.17% 15.75% Allocation $15.64 $13.59 $39.85 $15.17 $15.75 $100 Putting it all together
- 17. Where did that gift come from? #attribution 17 Mail Email Phone Email 2 Cookie Ad First Touch $100.00 – – – – Fractional Allocation $17.00 $18.60 $20.00 $20.90 $23.00 Weighted Allocation $15.64 $13.59 $39.85 $15.17 $15.75 Complete Dashboard
- 18. Where did that gift come from? #attribution 18 Reduce Mail Quality Maintain email, already contacting everybody Increase phone contacts Eliminate cookies Need to make investment decisions based on right combination of touchpoints. 52% 14% 33% 1% Mail Email Phone Cookie Ads Channel Attributed Revenue 66% 5% 20% 9% Mail Email Phone Cookie Ads Channel Expenditures What does all this mean?
- 19. Philadelphia Museum of Art / February 14, 2013 art institute of chicago case study
- 20. Where did that gift come from? #attribution March Lapsed Campaign Tested in Deep (60+ months) and Recent (12 months) Lapsed Segments. Split into the following segments: 1. Control (no contact) 2. Direct Mail 3. Email 4. Facebook Advertising 5. Cookie Matching 6. Telemarketing Bucket tests to see if lift is greater than the sum of their parts: 1. Direct Mail and Telemarketing 2. Direct Mail and Cookie Matching 20 Art Institute of Chicago Case Study
- 21. Where did that gift come from? #attribution Art Institute of Chicago Case Study
- 22. Where did that gift come from? #attribution 22 Art Institute of Chicago Case Study Direct mail
- 23. Where did that gift come from? #attribution 23 Art Institute of Chicago Case Study Short form Email template
- 24. Where did that gift come from? #attribution 24 Art Institute of Chicago Case Study Display and Facebook Ads
- 25. Where did that gift come from? #attribution • Withholding prospects during testing • Some organic donations will be made no matter what • Still need to look at campaign results “the old fashioned way” • Will need to update model yearly as communication preferences change • SPAM and digital privacy issues will make online communications tougher • Don’t know which people are particular channel responsive • Additional channels complicate the equation • Need a sophisticated data warehouse and tracking system in place • Ability to act on the information • What specific issues did AIC have to confront? 25 Considerations Art Institute of Chicago Case Study
- 26. email Where did that gift come from? #attribution Noise 26 tv phone facebook onsite twitter ads outdoor advertising search ads
- 27. Where did that gift come from? #attribution Best Practices for Multi-Channel Campaigns • Creative should match across channels • Offer should match • Fewer clicks the better • Contact people as many times as possible until diminishing returns (or opt-outs) • Analytics as part of planning process (not an afterthought) • Coding is Queen 27
- 28. Where did that gift come from? #attribution Evaluate This Session! Each entry is a chance to win an NTEN engraved iPad! or Online using #attribution at www.nten.org/ntc/eval Questions?? Angela Struebing The Lukens Company 703-845-8484 angela@thelukenscompany.com Amy Katherine Radick Art Institute of Chicago 312-443-3116 aradick@artic.edu