SlideShare a Scribd company logo
© 2016 Flashtalking. Confidential and may not be shared without permission.
25% 15% 30% 15% 15%
Keys to Successful Attribution
Steve Latham, Head of Analytics
May 2017
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 2
The customer journey is increasingly complex
Fractional Attribution
…Highlighting the need for accurate attribution
© 2016 Flashtalking. Confidential and may not be shared without permission.
><
The problem with “old” attribution
Last-touch (100%)
• Penalizes upper funnel ads
• Rewards low-funnel ads
• Encourages bad behavior (cookie bombing)
Even weighted
• Gives credit to every ad – regardless of impact
• Penalizes quality ads and rewards cookie bombing
First touch (100%)
• Rewards ads that may not have been noticed
• Penalizes middle and lower funnel
Time Decay
• Rewards most recent ads – regardless of impact
• Penalizes upper funnel ads
Display Imp: #1
Paid
Visit
#1
Display Imp: #4 Display Imp: #3
Display Imp: #2
Paid
Visit
#2
Non-
Paid
Visit
Display Imp: #5 Conversion
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 4
Bad attribution can do more harm than good
Fractional Attribution
30-50% of
Last-Touch
Signals are
WRONG
© 2016 Flashtalking. Confidential and may not be shared without permission.
><
Solution: algorithmic fractional attribution
Approach
• Unify customer journey (across platforms)
• Use validated, machine learning models to measure
contribution of each touchpoint
• Allocate fractional credit to each channel, placement,
creative and keyword
• Use those insights to optimize spend
Benefits:
• Accurate insights: cross-channel / full-funnel
• Transparency + confidence
• Higher returns from digital media
Display Imp: #1
Paid
Visit
#1
Display Imp: #4 Display Imp: #3
Display Imp: #2
Paid
Visit
#2
Non-
Paid
Visit
Display Imp: #5 Conversion
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 6
5 Keys to Successful Attribution
FRACTIONAL ATTRIBUTION
1. Unified data
2. Validated models
3. Intuitive reporting
4. Actionable recommendations
5. Ecosystem friendly
© 2016 Flashtalking. Confidential and may not be shared without permission.
><
7
Requirement #1: Comprehensive, unified data
FRACTIONAL ATTRIBUTION
Desktop
Impression
CONVERSION PATH ASSEMBLY
Non-Paid
Click
Conversion
Desktop Click
In-App
Impression
Desktop
Impression
In-App
Impression
Mobile Web
Impression
In-App
Impression
VISUALIZATION
DATA INPUTS
In-App Ads
Desktop Ads
Mobile Ads
Conversions
AD SERVER
ALGORITHMIC
MODELING
No redundant tagging.
No discrepancies.
Better data.
Better insights.
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 8
Requirement #2: Validated Models
FRACTIONAL ATTRIBUTION
Split the data into two representative groups: Training and Holdout
Run model(s) against the Training group
Validate the outputs against a holdout group to determine the predictive accuracy
Machine Learning Validation
Models must be validated. Validate predictions against a holdout group
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 9
Requirement #3: Intuitive reporting
FRACTIONAL ATTRIBUTION
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 10
Requirement #4: Actionable recommendations
FRACTIONAL ATTRIBUTION
© 2016 Flashtalking. Confidential and may not be shared without permission.
><
Fractional Attribution
11
FRACTIONAL ATTRIBUTION
Dynamic Ad Serving
Cookieless tracking
Delivery
Programmatic
Creative
Data and Reporting
Standard Ad Serving
Verification & Fraud
Cross-Device
Offline Sales
DMP Segments
Requirement #5: Ecosystem friendly (frictionless)
© 2016 Flashtalking. Confidential and may not be shared without permission.© 2016 Flashtalking. Confidential and may not be shared without permission.
…For those who thought Data and Creativity
could not co-exist…
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 13
Introducing Creative Analytics
Problem: traditional metrics are flawed
• Click-through rates are misleading
• Last-touch favors low funnel (penalizes upper funnel)
• Bias exists among placements and audiences
Solution: Creative Analytics
• Remove underperforming placements
• Cluster Users based on behavior
• Analyze Creatives within each Cluster
using algorithmic models
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 14
Inside Creative Analytics
1. Remove underperforming placements via attribution modeling
2. Cluster Users based on complexity and conversion probability (via Entropy and Turbulence modeling)
3. Analyze creatives within each cluster and Index based on probability of converting
Data Driven approach to measuring creative performance
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 15
Creative Case Study: Fortune 500 Bank
Last-touch: Dynamic ads had 15% higher CPA than Control (Static) group
With Creative Analytics we discovered:
• Dynamic outperformed by 40% (25% lower CPA)
• 38% of Control converters were exposed to Dynamic
Dynamic ads outperformed
Control group on each site
© 2016 Flashtalking. Confidential and may not be shared without permission.
>< 16
In Closing…
Multi-touch attribution is no longer a luxury – it’s a necessity
Benefits
• Better insights
• Superior performance
• Less friction
• Higher ROE: Return On Effort
Keys to Success
• Unified data
• Validated models
• Intuitive reporting
• Actionable recommendations
• Ecosystem friendly

More Related Content

What's hot

Festival of Marketing Datalicious OptimaHub Media Attribution
Festival of Marketing Datalicious OptimaHub Media AttributionFestival of Marketing Datalicious OptimaHub Media Attribution
Festival of Marketing Datalicious OptimaHub Media Attribution
Datalicious
 
ProgrammatiCon 2017 - The Future of Facebook Ads - Peter Podolinsky, ROI Hunter
ProgrammatiCon 2017 - The Future of Facebook Ads - Peter Podolinsky, ROI HunterProgrammatiCon 2017 - The Future of Facebook Ads - Peter Podolinsky, ROI Hunter
ProgrammatiCon 2017 - The Future of Facebook Ads - Peter Podolinsky, ROI Hunter
e-dialog GmbH
 
Hacking Marketing: Marketing Management in a Software World
Hacking Marketing: Marketing Management in a Software WorldHacking Marketing: Marketing Management in a Software World
Hacking Marketing: Marketing Management in a Software World
Ensighten
 
Mobile Apps: No Longer a Black Box of Digital Measurement
Mobile Apps: No Longer a Black Box of Digital MeasurementMobile Apps: No Longer a Black Box of Digital Measurement
Mobile Apps: No Longer a Black Box of Digital Measurement
Ensighten
 
Microsoft CRM xRM4Legal 2014 Spring Update
Microsoft CRM xRM4Legal 2014 Spring UpdateMicrosoft CRM xRM4Legal 2014 Spring Update
Microsoft CRM xRM4Legal 2014 Spring Update
David Blumentals
 
ProgrammatiCon 2017 - Driving Growth through Measurement - Emanuele Inforzato...
ProgrammatiCon 2017 - Driving Growth through Measurement - Emanuele Inforzato...ProgrammatiCon 2017 - Driving Growth through Measurement - Emanuele Inforzato...
ProgrammatiCon 2017 - Driving Growth through Measurement - Emanuele Inforzato...
e-dialog GmbH
 
OptimaHub MediaAttribution Presentation Deck
OptimaHub MediaAttribution Presentation DeckOptimaHub MediaAttribution Presentation Deck
OptimaHub MediaAttribution Presentation Deck
Datalicious
 
Jason Fischer, Conversant
Jason Fischer, ConversantJason Fischer, Conversant
Jason Fischer, Conversant
Digiday
 
Microsoft CRM xRM4legal 2015 Update
Microsoft CRM xRM4legal 2015 UpdateMicrosoft CRM xRM4legal 2015 Update
Microsoft CRM xRM4legal 2015 Update
David Blumentals
 
Wi-Fi Offload Summit - Monetise Thyself
Wi-Fi Offload Summit - Monetise ThyselfWi-Fi Offload Summit - Monetise Thyself
Wi-Fi Offload Summit - Monetise ThyselfAirTight Networks
 
Data-Driven Product Development
Data-Driven Product DevelopmentData-Driven Product Development
Data-Driven Product Development
Nitzan Achsaf
 
Master Cross-Channel Customer Recognition
Master Cross-Channel Customer RecognitionMaster Cross-Channel Customer Recognition
Master Cross-Channel Customer Recognition
Signal
 
Smarter Commerce for Insurance - 360 Degree Customer Experience Infographic
Smarter Commerce for Insurance - 360 Degree Customer Experience InfographicSmarter Commerce for Insurance - 360 Degree Customer Experience Infographic
Smarter Commerce for Insurance - 360 Degree Customer Experience Infographic
Lynn Kesterson-Townes
 
Digital marketing ROI - An introduction to attribution modelling
Digital marketing ROI - An introduction to attribution modellingDigital marketing ROI - An introduction to attribution modelling
Digital marketing ROI - An introduction to attribution modelling
Different Spin
 
The marketing automation manual part 1
The marketing automation manual part 1The marketing automation manual part 1
The marketing automation manual part 1
Tocquigny
 
An Overlooked Part of Outdoor Industry Retail? Data
An Overlooked Part of Outdoor Industry Retail? DataAn Overlooked Part of Outdoor Industry Retail? Data
An Overlooked Part of Outdoor Industry Retail? Data
GoSpotCheck
 
jeremy haft, viant
jeremy haft, viantjeremy haft, viant
jeremy haft, viant
Digiday
 
The Harvest Digital Guide to Attribution Modelling
The Harvest Digital Guide to Attribution ModellingThe Harvest Digital Guide to Attribution Modelling
The Harvest Digital Guide to Attribution Modelling
Mike Teasdale
 

What's hot (18)

Festival of Marketing Datalicious OptimaHub Media Attribution
Festival of Marketing Datalicious OptimaHub Media AttributionFestival of Marketing Datalicious OptimaHub Media Attribution
Festival of Marketing Datalicious OptimaHub Media Attribution
 
ProgrammatiCon 2017 - The Future of Facebook Ads - Peter Podolinsky, ROI Hunter
ProgrammatiCon 2017 - The Future of Facebook Ads - Peter Podolinsky, ROI HunterProgrammatiCon 2017 - The Future of Facebook Ads - Peter Podolinsky, ROI Hunter
ProgrammatiCon 2017 - The Future of Facebook Ads - Peter Podolinsky, ROI Hunter
 
Hacking Marketing: Marketing Management in a Software World
Hacking Marketing: Marketing Management in a Software WorldHacking Marketing: Marketing Management in a Software World
Hacking Marketing: Marketing Management in a Software World
 
Mobile Apps: No Longer a Black Box of Digital Measurement
Mobile Apps: No Longer a Black Box of Digital MeasurementMobile Apps: No Longer a Black Box of Digital Measurement
Mobile Apps: No Longer a Black Box of Digital Measurement
 
Microsoft CRM xRM4Legal 2014 Spring Update
Microsoft CRM xRM4Legal 2014 Spring UpdateMicrosoft CRM xRM4Legal 2014 Spring Update
Microsoft CRM xRM4Legal 2014 Spring Update
 
ProgrammatiCon 2017 - Driving Growth through Measurement - Emanuele Inforzato...
ProgrammatiCon 2017 - Driving Growth through Measurement - Emanuele Inforzato...ProgrammatiCon 2017 - Driving Growth through Measurement - Emanuele Inforzato...
ProgrammatiCon 2017 - Driving Growth through Measurement - Emanuele Inforzato...
 
OptimaHub MediaAttribution Presentation Deck
OptimaHub MediaAttribution Presentation DeckOptimaHub MediaAttribution Presentation Deck
OptimaHub MediaAttribution Presentation Deck
 
Jason Fischer, Conversant
Jason Fischer, ConversantJason Fischer, Conversant
Jason Fischer, Conversant
 
Microsoft CRM xRM4legal 2015 Update
Microsoft CRM xRM4legal 2015 UpdateMicrosoft CRM xRM4legal 2015 Update
Microsoft CRM xRM4legal 2015 Update
 
Wi-Fi Offload Summit - Monetise Thyself
Wi-Fi Offload Summit - Monetise ThyselfWi-Fi Offload Summit - Monetise Thyself
Wi-Fi Offload Summit - Monetise Thyself
 
Data-Driven Product Development
Data-Driven Product DevelopmentData-Driven Product Development
Data-Driven Product Development
 
Master Cross-Channel Customer Recognition
Master Cross-Channel Customer RecognitionMaster Cross-Channel Customer Recognition
Master Cross-Channel Customer Recognition
 
Smarter Commerce for Insurance - 360 Degree Customer Experience Infographic
Smarter Commerce for Insurance - 360 Degree Customer Experience InfographicSmarter Commerce for Insurance - 360 Degree Customer Experience Infographic
Smarter Commerce for Insurance - 360 Degree Customer Experience Infographic
 
Digital marketing ROI - An introduction to attribution modelling
Digital marketing ROI - An introduction to attribution modellingDigital marketing ROI - An introduction to attribution modelling
Digital marketing ROI - An introduction to attribution modelling
 
The marketing automation manual part 1
The marketing automation manual part 1The marketing automation manual part 1
The marketing automation manual part 1
 
An Overlooked Part of Outdoor Industry Retail? Data
An Overlooked Part of Outdoor Industry Retail? DataAn Overlooked Part of Outdoor Industry Retail? Data
An Overlooked Part of Outdoor Industry Retail? Data
 
jeremy haft, viant
jeremy haft, viantjeremy haft, viant
jeremy haft, viant
 
The Harvest Digital Guide to Attribution Modelling
The Harvest Digital Guide to Attribution ModellingThe Harvest Digital Guide to Attribution Modelling
The Harvest Digital Guide to Attribution Modelling
 

Similar to Deck 2: Attribution + Creative Analytics

AT Internet & Mazeberry: from analytics to a fully optimised marketing mix
AT Internet & Mazeberry: from analytics to a fully optimised marketing mixAT Internet & Mazeberry: from analytics to a fully optimised marketing mix
AT Internet & Mazeberry: from analytics to a fully optimised marketing mix
AT Internet
 
Investing Confidently (and Safely) in Programmatic
Investing Confidently (and Safely) in ProgrammaticInvesting Confidently (and Safely) in Programmatic
Investing Confidently (and Safely) in Programmatic
Encore Media Metrics
 
UNDERSTAND THE TRUE AUDIENCE FOR DIGITAL - Nielsen 2014
UNDERSTAND THE TRUE AUDIENCE FOR DIGITAL - Nielsen 2014UNDERSTAND THE TRUE AUDIENCE FOR DIGITAL - Nielsen 2014
UNDERSTAND THE TRUE AUDIENCE FOR DIGITAL - Nielsen 2014Karen Sanchez
 
How to improve your performance online - survey results Microsoft Dynamics Pa...
How to improve your performance online - survey results Microsoft Dynamics Pa...How to improve your performance online - survey results Microsoft Dynamics Pa...
How to improve your performance online - survey results Microsoft Dynamics Pa...
Patrick Dalle
 
Moderator - PerimeterX Webinar on Fraud
Moderator - PerimeterX Webinar on FraudModerator - PerimeterX Webinar on Fraud
Moderator - PerimeterX Webinar on FraudJack Chiang
 
230286802015PPT.pptx
230286802015PPT.pptx230286802015PPT.pptx
230286802015PPT.pptx
annalakshmi35
 
Tackle Ad Viewability in 2016
Tackle Ad Viewability in 2016Tackle Ad Viewability in 2016
Tackle Ad Viewability in 2016
Optimizely
 
Behavioral Analytics: Stop Guessing. Get Clarity - Priyanka Ved, Microsoft
Behavioral Analytics: Stop Guessing. Get Clarity - Priyanka Ved, MicrosoftBehavioral Analytics: Stop Guessing. Get Clarity - Priyanka Ved, Microsoft
Behavioral Analytics: Stop Guessing. Get Clarity - Priyanka Ved, Microsoft
DigiMarCon - Digital Marketing, Media and Advertising Conferences & Exhibitions
 
How to Benefit from Digital Distruption
How to Benefit from Digital DistruptionHow to Benefit from Digital Distruption
How to Benefit from Digital Distruption
Qualtrics
 
Digital Media Foundation
Digital Media Foundation Digital Media Foundation
Digital Media Foundation
shirlevangelisto
 
Adobe Summit - Data-Driven Marketing Attribution
Adobe Summit - Data-Driven Marketing AttributionAdobe Summit - Data-Driven Marketing Attribution
Adobe Summit - Data-Driven Marketing Attribution
Adometry by Google
 
Lake group media digital analytics platform
Lake group media   digital analytics platformLake group media   digital analytics platform
Lake group media digital analytics platform
kathystivaletti
 
Superweek - Marketing attribution after 2024
Superweek - Marketing attribution after 2024Superweek - Marketing attribution after 2024
Superweek - Marketing attribution after 2024
Jente De Ridder
 
New Adtaxi General Presentation - 9.28.16-2
New Adtaxi General Presentation - 9.28.16-2New Adtaxi General Presentation - 9.28.16-2
New Adtaxi General Presentation - 9.28.16-2Victoria Collins
 
Ad:Tech Data Summit - Sydney | 2013
Ad:Tech Data Summit - Sydney | 2013Ad:Tech Data Summit - Sydney | 2013
Ad:Tech Data Summit - Sydney | 2013
Louder
 
Wizaly
WizalyWizaly
Calculating the Effectiveness of Your Mobile Ad Campaigns
Calculating the Effectiveness of Your Mobile Ad CampaignsCalculating the Effectiveness of Your Mobile Ad Campaigns
Calculating the Effectiveness of Your Mobile Ad Campaigns
Appia
 
Multi-Channel Attribution - Where do Leads Come From?
Multi-Channel Attribution - Where do Leads Come From?Multi-Channel Attribution - Where do Leads Come From?
Multi-Channel Attribution - Where do Leads Come From?
Joshua Tree Internet Media, LLC
 
How To Promote Your Website
How To Promote Your WebsiteHow To Promote Your Website
How To Promote Your Website
Search Engine Pros
 
Getting Started with Online-to-Offline Measurement
Getting Started with Online-to-Offline MeasurementGetting Started with Online-to-Offline Measurement
Getting Started with Online-to-Offline Measurement
brandonwishnow
 

Similar to Deck 2: Attribution + Creative Analytics (20)

AT Internet & Mazeberry: from analytics to a fully optimised marketing mix
AT Internet & Mazeberry: from analytics to a fully optimised marketing mixAT Internet & Mazeberry: from analytics to a fully optimised marketing mix
AT Internet & Mazeberry: from analytics to a fully optimised marketing mix
 
Investing Confidently (and Safely) in Programmatic
Investing Confidently (and Safely) in ProgrammaticInvesting Confidently (and Safely) in Programmatic
Investing Confidently (and Safely) in Programmatic
 
UNDERSTAND THE TRUE AUDIENCE FOR DIGITAL - Nielsen 2014
UNDERSTAND THE TRUE AUDIENCE FOR DIGITAL - Nielsen 2014UNDERSTAND THE TRUE AUDIENCE FOR DIGITAL - Nielsen 2014
UNDERSTAND THE TRUE AUDIENCE FOR DIGITAL - Nielsen 2014
 
How to improve your performance online - survey results Microsoft Dynamics Pa...
How to improve your performance online - survey results Microsoft Dynamics Pa...How to improve your performance online - survey results Microsoft Dynamics Pa...
How to improve your performance online - survey results Microsoft Dynamics Pa...
 
Moderator - PerimeterX Webinar on Fraud
Moderator - PerimeterX Webinar on FraudModerator - PerimeterX Webinar on Fraud
Moderator - PerimeterX Webinar on Fraud
 
230286802015PPT.pptx
230286802015PPT.pptx230286802015PPT.pptx
230286802015PPT.pptx
 
Tackle Ad Viewability in 2016
Tackle Ad Viewability in 2016Tackle Ad Viewability in 2016
Tackle Ad Viewability in 2016
 
Behavioral Analytics: Stop Guessing. Get Clarity - Priyanka Ved, Microsoft
Behavioral Analytics: Stop Guessing. Get Clarity - Priyanka Ved, MicrosoftBehavioral Analytics: Stop Guessing. Get Clarity - Priyanka Ved, Microsoft
Behavioral Analytics: Stop Guessing. Get Clarity - Priyanka Ved, Microsoft
 
How to Benefit from Digital Distruption
How to Benefit from Digital DistruptionHow to Benefit from Digital Distruption
How to Benefit from Digital Distruption
 
Digital Media Foundation
Digital Media Foundation Digital Media Foundation
Digital Media Foundation
 
Adobe Summit - Data-Driven Marketing Attribution
Adobe Summit - Data-Driven Marketing AttributionAdobe Summit - Data-Driven Marketing Attribution
Adobe Summit - Data-Driven Marketing Attribution
 
Lake group media digital analytics platform
Lake group media   digital analytics platformLake group media   digital analytics platform
Lake group media digital analytics platform
 
Superweek - Marketing attribution after 2024
Superweek - Marketing attribution after 2024Superweek - Marketing attribution after 2024
Superweek - Marketing attribution after 2024
 
New Adtaxi General Presentation - 9.28.16-2
New Adtaxi General Presentation - 9.28.16-2New Adtaxi General Presentation - 9.28.16-2
New Adtaxi General Presentation - 9.28.16-2
 
Ad:Tech Data Summit - Sydney | 2013
Ad:Tech Data Summit - Sydney | 2013Ad:Tech Data Summit - Sydney | 2013
Ad:Tech Data Summit - Sydney | 2013
 
Wizaly
WizalyWizaly
Wizaly
 
Calculating the Effectiveness of Your Mobile Ad Campaigns
Calculating the Effectiveness of Your Mobile Ad CampaignsCalculating the Effectiveness of Your Mobile Ad Campaigns
Calculating the Effectiveness of Your Mobile Ad Campaigns
 
Multi-Channel Attribution - Where do Leads Come From?
Multi-Channel Attribution - Where do Leads Come From?Multi-Channel Attribution - Where do Leads Come From?
Multi-Channel Attribution - Where do Leads Come From?
 
How To Promote Your Website
How To Promote Your WebsiteHow To Promote Your Website
How To Promote Your Website
 
Getting Started with Online-to-Offline Measurement
Getting Started with Online-to-Offline MeasurementGetting Started with Online-to-Offline Measurement
Getting Started with Online-to-Offline Measurement
 

Recently uploaded

The Building Blocks of QuestDB, a Time Series Database
The Building Blocks of QuestDB, a Time Series DatabaseThe Building Blocks of QuestDB, a Time Series Database
The Building Blocks of QuestDB, a Time Series Database
javier ramirez
 
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
ewymefz
 
Everything you wanted to know about LIHTC
Everything you wanted to know about LIHTCEverything you wanted to know about LIHTC
Everything you wanted to know about LIHTC
Roger Valdez
 
一比一原版(Coventry毕业证书)考文垂大学毕业证如何办理
一比一原版(Coventry毕业证书)考文垂大学毕业证如何办理一比一原版(Coventry毕业证书)考文垂大学毕业证如何办理
一比一原版(Coventry毕业证书)考文垂大学毕业证如何办理
74nqk8xf
 
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
John Andrews
 
一比一原版(UPenn毕业证)宾夕法尼亚大学毕业证成绩单
一比一原版(UPenn毕业证)宾夕法尼亚大学毕业证成绩单一比一原版(UPenn毕业证)宾夕法尼亚大学毕业证成绩单
一比一原版(UPenn毕业证)宾夕法尼亚大学毕业证成绩单
ewymefz
 
The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...
jerlynmaetalle
 
一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单
ewymefz
 
一比一原版(UniSA毕业证书)南澳大学毕业证如何办理
一比一原版(UniSA毕业证书)南澳大学毕业证如何办理一比一原版(UniSA毕业证书)南澳大学毕业证如何办理
一比一原版(UniSA毕业证书)南澳大学毕业证如何办理
slg6lamcq
 
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Subhajit Sahu
 
Machine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptxMachine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptx
balafet
 
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
ahzuo
 
06-04-2024 - NYC Tech Week - Discussion on Vector Databases, Unstructured Dat...
06-04-2024 - NYC Tech Week - Discussion on Vector Databases, Unstructured Dat...06-04-2024 - NYC Tech Week - Discussion on Vector Databases, Unstructured Dat...
06-04-2024 - NYC Tech Week - Discussion on Vector Databases, Unstructured Dat...
Timothy Spann
 
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
pchutichetpong
 
原版制作(swinburne毕业证书)斯威本科技大学毕业证毕业完成信一模一样
原版制作(swinburne毕业证书)斯威本科技大学毕业证毕业完成信一模一样原版制作(swinburne毕业证书)斯威本科技大学毕业证毕业完成信一模一样
原版制作(swinburne毕业证书)斯威本科技大学毕业证毕业完成信一模一样
u86oixdj
 
Q1’2024 Update: MYCI’s Leap Year Rebound
Q1’2024 Update: MYCI’s Leap Year ReboundQ1’2024 Update: MYCI’s Leap Year Rebound
Q1’2024 Update: MYCI’s Leap Year Rebound
Oppotus
 
一比一原版(UIUC毕业证)伊利诺伊大学|厄巴纳-香槟分校毕业证如何办理
一比一原版(UIUC毕业证)伊利诺伊大学|厄巴纳-香槟分校毕业证如何办理一比一原版(UIUC毕业证)伊利诺伊大学|厄巴纳-香槟分校毕业证如何办理
一比一原版(UIUC毕业证)伊利诺伊大学|厄巴纳-香槟分校毕业证如何办理
ahzuo
 
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
2023240532
 
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
AbhimanyuSinha9
 
一比一原版(爱大毕业证书)爱丁堡大学毕业证如何办理
一比一原版(爱大毕业证书)爱丁堡大学毕业证如何办理一比一原版(爱大毕业证书)爱丁堡大学毕业证如何办理
一比一原版(爱大毕业证书)爱丁堡大学毕业证如何办理
g4dpvqap0
 

Recently uploaded (20)

The Building Blocks of QuestDB, a Time Series Database
The Building Blocks of QuestDB, a Time Series DatabaseThe Building Blocks of QuestDB, a Time Series Database
The Building Blocks of QuestDB, a Time Series Database
 
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
 
Everything you wanted to know about LIHTC
Everything you wanted to know about LIHTCEverything you wanted to know about LIHTC
Everything you wanted to know about LIHTC
 
一比一原版(Coventry毕业证书)考文垂大学毕业证如何办理
一比一原版(Coventry毕业证书)考文垂大学毕业证如何办理一比一原版(Coventry毕业证书)考文垂大学毕业证如何办理
一比一原版(Coventry毕业证书)考文垂大学毕业证如何办理
 
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
 
一比一原版(UPenn毕业证)宾夕法尼亚大学毕业证成绩单
一比一原版(UPenn毕业证)宾夕法尼亚大学毕业证成绩单一比一原版(UPenn毕业证)宾夕法尼亚大学毕业证成绩单
一比一原版(UPenn毕业证)宾夕法尼亚大学毕业证成绩单
 
The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...
 
一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单
 
一比一原版(UniSA毕业证书)南澳大学毕业证如何办理
一比一原版(UniSA毕业证书)南澳大学毕业证如何办理一比一原版(UniSA毕业证书)南澳大学毕业证如何办理
一比一原版(UniSA毕业证书)南澳大学毕业证如何办理
 
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
 
Machine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptxMachine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptx
 
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
 
06-04-2024 - NYC Tech Week - Discussion on Vector Databases, Unstructured Dat...
06-04-2024 - NYC Tech Week - Discussion on Vector Databases, Unstructured Dat...06-04-2024 - NYC Tech Week - Discussion on Vector Databases, Unstructured Dat...
06-04-2024 - NYC Tech Week - Discussion on Vector Databases, Unstructured Dat...
 
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
 
原版制作(swinburne毕业证书)斯威本科技大学毕业证毕业完成信一模一样
原版制作(swinburne毕业证书)斯威本科技大学毕业证毕业完成信一模一样原版制作(swinburne毕业证书)斯威本科技大学毕业证毕业完成信一模一样
原版制作(swinburne毕业证书)斯威本科技大学毕业证毕业完成信一模一样
 
Q1’2024 Update: MYCI’s Leap Year Rebound
Q1’2024 Update: MYCI’s Leap Year ReboundQ1’2024 Update: MYCI’s Leap Year Rebound
Q1’2024 Update: MYCI’s Leap Year Rebound
 
一比一原版(UIUC毕业证)伊利诺伊大学|厄巴纳-香槟分校毕业证如何办理
一比一原版(UIUC毕业证)伊利诺伊大学|厄巴纳-香槟分校毕业证如何办理一比一原版(UIUC毕业证)伊利诺伊大学|厄巴纳-香槟分校毕业证如何办理
一比一原版(UIUC毕业证)伊利诺伊大学|厄巴纳-香槟分校毕业证如何办理
 
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
 
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
 
一比一原版(爱大毕业证书)爱丁堡大学毕业证如何办理
一比一原版(爱大毕业证书)爱丁堡大学毕业证如何办理一比一原版(爱大毕业证书)爱丁堡大学毕业证如何办理
一比一原版(爱大毕业证书)爱丁堡大学毕业证如何办理
 

Deck 2: Attribution + Creative Analytics

  • 1. © 2016 Flashtalking. Confidential and may not be shared without permission. 25% 15% 30% 15% 15% Keys to Successful Attribution Steve Latham, Head of Analytics May 2017
  • 2. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 2 The customer journey is increasingly complex Fractional Attribution …Highlighting the need for accurate attribution
  • 3. © 2016 Flashtalking. Confidential and may not be shared without permission. >< The problem with “old” attribution Last-touch (100%) • Penalizes upper funnel ads • Rewards low-funnel ads • Encourages bad behavior (cookie bombing) Even weighted • Gives credit to every ad – regardless of impact • Penalizes quality ads and rewards cookie bombing First touch (100%) • Rewards ads that may not have been noticed • Penalizes middle and lower funnel Time Decay • Rewards most recent ads – regardless of impact • Penalizes upper funnel ads Display Imp: #1 Paid Visit #1 Display Imp: #4 Display Imp: #3 Display Imp: #2 Paid Visit #2 Non- Paid Visit Display Imp: #5 Conversion
  • 4. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 4 Bad attribution can do more harm than good Fractional Attribution 30-50% of Last-Touch Signals are WRONG
  • 5. © 2016 Flashtalking. Confidential and may not be shared without permission. >< Solution: algorithmic fractional attribution Approach • Unify customer journey (across platforms) • Use validated, machine learning models to measure contribution of each touchpoint • Allocate fractional credit to each channel, placement, creative and keyword • Use those insights to optimize spend Benefits: • Accurate insights: cross-channel / full-funnel • Transparency + confidence • Higher returns from digital media Display Imp: #1 Paid Visit #1 Display Imp: #4 Display Imp: #3 Display Imp: #2 Paid Visit #2 Non- Paid Visit Display Imp: #5 Conversion
  • 6. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 6 5 Keys to Successful Attribution FRACTIONAL ATTRIBUTION 1. Unified data 2. Validated models 3. Intuitive reporting 4. Actionable recommendations 5. Ecosystem friendly
  • 7. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 7 Requirement #1: Comprehensive, unified data FRACTIONAL ATTRIBUTION Desktop Impression CONVERSION PATH ASSEMBLY Non-Paid Click Conversion Desktop Click In-App Impression Desktop Impression In-App Impression Mobile Web Impression In-App Impression VISUALIZATION DATA INPUTS In-App Ads Desktop Ads Mobile Ads Conversions AD SERVER ALGORITHMIC MODELING No redundant tagging. No discrepancies. Better data. Better insights.
  • 8. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 8 Requirement #2: Validated Models FRACTIONAL ATTRIBUTION Split the data into two representative groups: Training and Holdout Run model(s) against the Training group Validate the outputs against a holdout group to determine the predictive accuracy Machine Learning Validation Models must be validated. Validate predictions against a holdout group
  • 9. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 9 Requirement #3: Intuitive reporting FRACTIONAL ATTRIBUTION
  • 10. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 10 Requirement #4: Actionable recommendations FRACTIONAL ATTRIBUTION
  • 11. © 2016 Flashtalking. Confidential and may not be shared without permission. >< Fractional Attribution 11 FRACTIONAL ATTRIBUTION Dynamic Ad Serving Cookieless tracking Delivery Programmatic Creative Data and Reporting Standard Ad Serving Verification & Fraud Cross-Device Offline Sales DMP Segments Requirement #5: Ecosystem friendly (frictionless)
  • 12. © 2016 Flashtalking. Confidential and may not be shared without permission.© 2016 Flashtalking. Confidential and may not be shared without permission. …For those who thought Data and Creativity could not co-exist…
  • 13. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 13 Introducing Creative Analytics Problem: traditional metrics are flawed • Click-through rates are misleading • Last-touch favors low funnel (penalizes upper funnel) • Bias exists among placements and audiences Solution: Creative Analytics • Remove underperforming placements • Cluster Users based on behavior • Analyze Creatives within each Cluster using algorithmic models
  • 14. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 14 Inside Creative Analytics 1. Remove underperforming placements via attribution modeling 2. Cluster Users based on complexity and conversion probability (via Entropy and Turbulence modeling) 3. Analyze creatives within each cluster and Index based on probability of converting Data Driven approach to measuring creative performance
  • 15. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 15 Creative Case Study: Fortune 500 Bank Last-touch: Dynamic ads had 15% higher CPA than Control (Static) group With Creative Analytics we discovered: • Dynamic outperformed by 40% (25% lower CPA) • 38% of Control converters were exposed to Dynamic Dynamic ads outperformed Control group on each site
  • 16. © 2016 Flashtalking. Confidential and may not be shared without permission. >< 16 In Closing… Multi-touch attribution is no longer a luxury – it’s a necessity Benefits • Better insights • Superior performance • Less friction • Higher ROE: Return On Effort Keys to Success • Unified data • Validated models • Intuitive reporting • Actionable recommendations • Ecosystem friendly