Ce diaporama a bien été signalé.
Le téléchargement de votre SlideShare est en cours. ×

Attribution 101

Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Publicité
Chargement dans…3
×

Consultez-les par la suite

1 sur 41 Publicité

Plus De Contenu Connexe

Diaporamas pour vous (20)

Publicité

Similaire à Attribution 101 (20)

Plus par InnoTech (20)

Publicité

Attribution 101

  1. 1. Attribution 101 Giving Credit Where Credit Is Due Aruna Thota, Director of Product Management, Adometry November 8, 2012
  2. 2. The Attribution Problem © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 2
  3. 3. Timberland © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 3
  4. 4. Timberland •Who gets the credit? •? © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 4
  5. 5. What Is Attribution? •Attribution is a way to assign value to each channel or marketing touch point that played a role in influencing the consumer to make a purchase. •Currently Attribution focus is mostly on online media © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 5
  6. 6. Types Of Attribution Models © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 6
  7. 7. Current State Of Attribution Sources: 6. Digital Attribution Comes Of Age © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 7
  8. 8. But, Wait… Total Revenue = Offline Revenue + Online Revenue © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 8
  9. 9. And…. Advertising Revenue Offline Advertising Offline Revenue Online Advertising Online Revenue © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 9
  10. 10. There Is More… © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 10
  11. 11. Types Of Attribution Advertising Revenue Offline Advertising Offline Revenue Online Advertising Online Revenue 1. Digital Attribution 4. Multi Device © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 11
  12. 12. Why Is Attribution Important? © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 12
  13. 13. Marketers Want To Be Data Driven •91% of high level decision makers believe successful brands use data to drive decisions •When it comes to CMO’s 100% believe data and measurement is key to driving decisions Source: http://www.iab.net/media/file/2012-BRITE-NYAMA-Marketing-ROI-Study.pdf © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 13
  14. 14. But, Marketers Struggle With Data Source: http://www.iab.net/media/file/2012-BRITE-NYAMA-Marketing-ROI-Study.pdf © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 14
  15. 15. And They Fail To Measure ROI •57% do not tie marketing budgets to ROI •37% don’t include financial metrics in measurement of ROI •82% don’t include both spend and returns to evaluate ROI •22% use brand awareness as the sole measure for marketing •Wrong metrics  Wrong marketing investments Source: http://www.iab.net/media/file/2012-BRITE-NYAMA-Marketing-ROI-Study.pdf © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 15
  16. 16. Benefits Of Attribution © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 16
  17. 17. Other Benefits 1. Understand true CPA’s and negotiate CPA with Affiliates 2. Reduce and eventually eliminate wasted spend 3. Understand non-linear performance of marketing 4. Improve planning © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 17
  18. 18. Example Using Display Impact On Search © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 18
  19. 19. Display Refines Search Queries Exposed to Display Ad No Display Exposure  More generic search terms 5 – 25% more • More expensive relevant search • More long tail keywords queries1 • Increased PPC complexity Sources 1.Display Advertising Impact: Search Lift and Social Influence (A field experiment by researchers at Stanford university) © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 19
  20. 20. Other Studies3 Indicate Positive Lifts Across Multiple Kpis • For consumers exposed to brand display campaigns 1 in 5 1 in 3 50% 10% Conduct Visit More More related brands Time on Money search web site site Online Higher More Higher Better Order Loyal Repeat Advocates Value $$ $$$ $$$$ $ Sources: 3. The Silent Click: Building Brands Online © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 20
  21. 21. Startup Steps For Attribution © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 21
  22. 22. What Level Of Attribution Do You Need? •Depends on the complexity of your business – start by analyzing your referrer data © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 22
  23. 23. Identify The Right Metrics •If you start measuring something you start optimizing it, even if it is wrong – example CTR © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 23
  24. 24. Test Things To Gain Confidence •Start with digital first © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 24
  25. 25. Bring Everyone Onboard •Change Organizational Structures and Incentive Mechanisms © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 25
  26. 26. The Adometry Attribution Solution © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 26
  27. 27. Data Driven Attribution •Adometry uses a bottom-up approach to attribution by first assigning fractional credit at the most granular impression level. •Allow the data to point out the hidden correlations and insights within your marketing efforts. © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 27
  28. 28. The Adometry Attribution Methodology © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 28
  29. 29. The Attribution Problem Looking backward, a user had events E1, E2, E3, E4 & then converted. E1 E2 E3 E4 $ What fractional credit goes to each Ei? © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 29
  30. 30. The Attribution Problem Looking backward, a user had events E1, E2, E3, E4 & then converted. E1 E2 E3 E4 Conversion Event What fractional credit goes to each Ei? © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 30
  31. 31. Adometry’s Advanced Attribution How important is each event in driving conversions? E1 E2 E3 E4 % Conversion E1 E2 E3 E4 % Conversion © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 31
  32. 32. What Is Optimization? Making incremental changes to media spend, to maximize lift in a target KPI Across channels Search Display Social email Offline Within channels Site/Placements © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 32
  33. 33. Average vs Incremental Metrics Marginal eCPA ≠ Average eCPA Optimization requires a non-linear model © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 33
  34. 34. Superior Advanced Optimization Non-Linear Model Constraint Driven More Accurate Starting Point Optimization Results Planned Budget Optimal Spend Allocations Reallocation Max Projected Constraints KPI Lift Training Data Exogenous data Daily ad spend KPI’s • Economic Indices • # Conversions • Display • Social • Weather • # Unique visitors • Search • TV • Custom • $ Revenue/Profit • Email • Offline • Affiliate • Custom © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 34 34
  35. 35. Advanced Attribution Input Automation from Media Partners Ad & Audience Verification Output Automatically to DSPs Viewability & Search Management Exclusive Reach Frequency Capping Funnel Stage Analysis Media Modeler Report Labs Comprehensive, Data-Driven Methodology | End-to-End Service | Ease of Integration © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 35
  36. 36. Advanced Attribution Viewability & Brand Safety: • Am I getting what I paid for? • Is the target audience seeing my ad? • Are my ads placed on brand damaging sites? MONITOR CAMPAIGNS TO MITIGATE WASTE • Ensure viewability • Target the right audience • Protect your brand IMPACT: Immediate 10-20% waste reduction © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 36
  37. 37. Advanced Attribution Multiple Channels + Multiple Campaigns = More Work than Time: • How do I get more out of my ad spend? • Where do I start to make adjustments? IMPROVE PERFORMANCE AND PREDICT “Easy Button” to SUCCESS optimize campaign performance based on constraints and key performance indicators IMPACT: Immediate 20-40% aggregate ROAS lift through optimization © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 37
  38. 38. Advanced Attribution Making Decisions Based on the Wrong Metrics: • How do I know what is really influencing customer conversions? • How are the channels interacting with each other? GLEAN INSIGHTS TO BUILD BETTER • Determine what is CAMPAIGNS and isn’t working • Cycle the improvements IMPACT: Longer term 25-75% aggregate ROAS lift through better campaign development & strategies through attribution © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 38
  39. 39. Selecting An Attribution Vendor © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 39
  40. 40. What To Look For In An Attribution Solution • Quality & completeness of data • Data driven & self learning model • Flexible and customizable • Insights & recommendations • “Easy button” • Best practices guidance • Vendor as a partner © Copyright 2012 Adometry, Inc. All Rights Reserved. Confidential. 40
  41. 41. Thank you Aruna Thota, Director of Product Management, Adometry aruna.thota@adometry.com

Notes de l'éditeur

  • Key FindingsForrester’s study yielded six key findings:Practitioners of attribution indicate general consensus on the definitions, benefits, and expectations of attribution.Forty-four percent of interactive marketers don’t have processes in place to assign credit to their efforts, not even rudimentary attribution such as “last event,” indicating that adoption of attribution has room to grow.Few marketers, agencies, or publishers stand behind last-event attribution, but it’s still commonly used as a standard for attribution.Algorithmic attribution models are gaining acceptance, but some in the market remain skeptical.Publishers are cautiously enthusiastic about marketers using attribution to show the value of upper-funnel inventory.Attribution across devices represents a major challenge.
  • In additionDisplay has an average conversion lag that is 6 times larger than searchDisplay when preceding other channels shortens the conversion lagDisplay is the introducer in almost 80% of the conversions it participates inIn visitor paths towards conversion display is one of the first few media touch points and consistently lifts other lower funnel media such as paid, organic search and Affiliates
  • Although there are several anecdotal examples that say 90% of users exposed to display ads are not engaged, that is the danger we encounter when we use flawed metrics such as click through rates to measure impact of display. When there are formalized A/B tests and experiments structured to measure display, we see a different story emerging that shows increased search relevance, increased site visits, increase in time on page and larger order sizes which all ties back to hard revenue numbers for the advertisers.
  • To understand the weight or credit due to an event in a particular sequence, we first find all examples of that sequence – both converting and non-converting.  From this we can calculate the conversion rate for that sequence.Next, we find a similar sequence of events that excludes one of the events from the prior sequence – for example, the first event.  Again, we find both converting and non-converting examples.  We then calculate the conversion rate for this (second) sequence.Now, by comparing the conversion rates of the sequences we can determine the weight or impact of the the missing event.  If the rates are identical or similar then the missing event deserves little or no credit whereas if the rates are different then that missing event had an impact and deserves credit.This is an obviously simplified explanation. The complexity is in the details.Number of sequence and length of sequences will varyWe normalize the weights such that they add up to one (1)Tons of events - "big data problem"If there is only one event in the sequence it will get 100% of the creditAn event is actually a complicated thing – has lots of attributes – site, campaign, placement, creative concept, tactic, etc.   we calculate at the lowest level and roll the results upAlso frequency and recency.But this is the stuff that the engineers have all worked out in a principled manner.  Machine learning, etc.
  • Moving from interesting to ACTIONABLE!
  • Attribution is sync’ed with the lifecycle of a campaign- Monitor Campaigns to mitigate waste – ad and audience verification to ensure viewability and target the right audience; Impact: recognize an immediate 10-20% waste reduction through elimination of out-of-compliance and non-viewed display ads - Improve Performance and predict success – automate the numerous adjustments that can drive significant improvement in ROAS and optimize in-flight campaigns. The complexity of cross-media interactions is too challenging to be analyzed and managed manually; it requires sophisticated technology to automate the process and make recommendations for media buyers and planners - an "Easy Button" to optimize campaign performance based on constraints and key performance indicators. Impact: Immediate 20–40% aggregate ROAS lift through optimization - Glean Insights to build better campaigns – looks at the entire stream of media to evaluate and measure campaigns to determine what is and isn't working. Impact: Longer term 25-75% aggregate ROAS lift through better campaign development and strategies through attribution
  • - Monitor Campaigns to mitigate waste – ad and audience verification to ensure viewability and target the right audience; Impact: recognize an immediate 10-20% waste reduction through elimination of out-of-compliance and non-viewed display ads - Improve Performance and predict success – automate the numerous adjustments that can drive significant improvement in ROAS and optimize in-flight campaigns. The complexity of cross-media interactions is too challenging to be analyzed and managed manually; it requires sophisticated technology to automate the process and make recommendations for media buyers and planners - an "Easy Button" to optimize campaign performance based on constraints and key performance indicators. Impact: Immediate 20–40% aggregate ROAS lift through optimization - Glean Insights to build better campaigns – looks at the entire stream of media to evaluate and measure campaigns to determine what is and isn't working. Impact: Longer term 25-75% aggregate ROAS lift through better campaign development and strategies through attribution

×