2. 1. What is Multi-Touch Attribution (Simply Put)
2. What MMA SMoX Research Proves about MTA
3. The Fundamental, Case for MTA
4. MTA is easy right? Nope!
5. Finally, What is the MMA Doing About it
Want a copy, send me an email: greg@mmaglobal.com
What to Do Today
3. Who We Are: MMA Global Board of Directors
Ernesto Echeverri
LATAM Regional Rep
Dir. Mktg USA,
Canada & Caribbean
John Costello
Global Chair Emeritus
Former Pres., Global
Mktg & Innovation
Luis Di Como
Exec Comm at Large
SVP, Global Media
Sanjay Gupta
Chair
CMO
Susan Canavari
Chief Brand Officer
Marc Mathieu
CMO
Peter Hamilton
CEO
Kellyn Kenny
VP,
Marketing
Jeff Lucas
CRO
Peter McGuinness
Chief Mktg &
Brand Officer
Ilonka Laviz
Global Digital Mktg Dir.,
Global eBusiness
Dave Morgan
CEO
Will Kassoy
CEO
Tim Mahoney
CMO
Andrew Sherrard
CMO
Tom Kenney
CEO
Karin Timpone
Global Mktg Officer
Alberto ‘Banano’
Pardo
LATAM Regional Rep
Founder & CEO
John Trimble
CRO
Michael Donnelly
NA Regional Rep
SVP, Global
Digital Marketing
David Lowes
EMEA Regional Rep
CMO
Allan Thygesen
President,
Americas
3
Regional
Louis Paskalis
NA Regional Rep.
SVP Enterprise Media
Executive Bank Of
America
Carolyn Everson
Global Secretary
VP, Global Mktg
Solutions
Stephen
McCarthy
Global Treasurer
CFO
(xAd)
Jack Philbin
Exec Comm at Large
Co-Founder
& CEO
Cameron Clayton
CEO & General
Manager
Greg Stuart
MMA President
CEO
Wanda Young
CMO
4. Fundamentally, a Trade Group is all About
Looking into the future. Then Driving everyone
there.
4
5. First, What is MTA
What is it, what is it
not, Why does it
matter?
5
Greg Stuart
greg@mmaglobal.com
+1 631 702 0682
6. Fundamentally, SMoX is MTA, which is…
6
Multi-Touch Attribution: The science of
using advanced analytics,
• on user level data,
• to allocate proportional credit,
• across a granular list of marketing
touchpoints across many,
• and hopefully all, online and offline channels,
leading to a desired customer outcome.
Excluded: Traditional MMM, brand tracking and last-touch attribution methods
7. What, How & Why of MTA
7
1a: WHAT Multi-Touch Attribution IS
I. Documented Definition: Science of using advanced analytics on user level data to
allocate proportional credit across a granular list of marketing touch points across
many and hopefully all online and off-line channels, leading to a desired customer
outcome.
II. Or put another way: MTA is an analytical method to distinguish what's working when
dozens of marketing activities and approaches are occurring at the same time in order
to estimate what is driving incremental lift.
i. Or, to measure impact so as to do more of what is working and less of what is not,
therefore, increasing marketing productivity.
CONFIDENTIAL
8. 8
1b: WHAT Multi-Touch Attribution IS NOT
I. Does not include methods that exclusively use Media Mix Modeling (MMM).
I. Also, is not brand/campaign tracking, or pop-up type surveys a result of ad exposure.
II. Is NOT limited to measurement of only whole media channels (TV, Radio, Digital, etc.)
in aggregate as sufficient.
III. It is NOT last touch measurement.
IV. Is NOT retrospective, requiring 2-3 years of back data to make decisions on today’s
media allocations.
V. Is NOT restricted to paid advertising as it can include the effect of owned media page
views and even social media under certain circumstances.
What, How & Why of MTA
CONFIDENTIAL
9. In 2011, Mobile Had
a Real Issue
To Solve that the MMA
raised $3 Million
to Gain an
Understanding of the
Value/ROI of Mobile in
a Marketing Mix
9
Greg Stuart
greg@mmaglobal.com
+1 631 702 0682
10. SMoX Was Our Answer – It is MTA
So far We Have Measured 11 In-Market Campaigns
10
Association of National
Advertisers
American Association of
Advertising Agencies
Partners
2
12. SMoX Historical Patterns in Optimized Levels of Mobile
in the Marketing Mix
12
1%
5% 4%
6% 5%
12% 11%
16%
26%
20%
15%
12%
15% 15%
19%
17% 17%
33%
0%
5%
10%
15%
20%
25%
30%
35%
AT&T Gold Peak
Tea/Coca Cola
MasterCard Walmart 825 Coca Cola
China
Coca Cola
Brazil
Unilever
Magnum
Allstate Wendys
% of mobile display in the mx (ex search)
Optimal allocations based on SMoX Studies
considering relative ROA/efficiency vs other media
Gap to Optimal
Actual Spend
To Come
13. SMoX Historical ROI Patterns when Mobile is Optimized
in the Mix and when Mobile is Optimized in itself
13
18%
7%
17% 14% 16%
60%
13% 7%
166%
0%
25%
50%
75%
100%
125%
150%
175%
AT&T Gold Peak
Tea/Coca
Cola
MasterCard Walmart 825 Coca Cola
China
Coca Cola
Brazil
Unilever
Magnum
Allstate Wendys
% of mobile display in the mx (ex search)
Increase in Key Business Outcomes
Based Mobile in the Mix vs. No Mobile
Increase in Key
Business Outcomes
To Come
Sales
SalesSalesSalesSalesBrand
Sales
14. But SMoX Informed
the Case for MTA
What Did SMoX Tell us
about the Opportunity
for MTA and it’s
importance
14
Greg Stuart
greg@mmaglobal.com
+1 631 702 0682
15. 15
Just Looking at Ads, Doesn’t tell us the Whole Story
15
Branding:
Aided Awareness/
$ spent*
Sales:
ROI
Mobile Display
with weather
targeting
Mobile
Display
Campaign Average
(across all media) 100 100
*Index is based on Number of people who became aware of Magnum per $ spent.
200 147
175 ZERO
16. *Index is based on Number of people who became more likely to consider Allstate / $ spent
We’ve Consistently Seen How Targeting Improves
Performance. But how much?
Mobile Video Targeting Consideration / $ Spent
+ Behavioral
targeting
+ Contextual
targeting
Demographic
targeting 100
16
320
191
17. And New Data integrations makes new
Opportunities (audience & performance) possible
Foot traffic /
$ spent*
Commuter
Coupon User
Campaign Average
(across all media) 100
518
500
17
18. …How else do you assess the value of ad unit combination #359:
Or, Mobile Audio channel targeting men, who
have visited a dealership in the last 7 days,
between 11a-1p weekdays via Pandora
…Versus the 358 other ad combinations
Based on That, The Case for MTA is…:
18
20. Translating Cost of MTA to Revenue: Example
Actual Optimized
Impact from
optimization
Campaign Budget $15 M $15 M
Estimated Revenue due to Media $30 M $34 M + $4 M
Attribution Payback (incremental
revenue / cost of research) :
$20 : $1
Incremental revenue if applied company
wide (assume total advertising spend of
$500 MM):
$131 MM
Based on data from a real SMoX campaign, manipulated to hide actuals but to maintain directional conclusions
21. MTA is IT! Easy,
Right?
Well, not so much.
21
Greg Stuart
greg@mmaglobal.com
+1 631 702 0682
22. State of MTA Adoption: 1/3rd of marketers currently
use MTA. 75% will be using in 18 months.
22
34%
76%
15%
17%
10%
25%
0%
20%
40%
60%
80%
100%
Yes we curently use
MTA
We will use in 6 months We will use in 12
months
We will use in 18
months
We dont plan to use in
the near future
75%
Does your company currently use a multi touch attribution (MTA) solution or do you plan to use one in the future?
N=412, Total Sample
23. But that Same Research Showed Huge
Dissatisfaction with MTA
Confidential: Cannot be shared without permission from the Mobile Marketing Association
MMA interviews and a quantitative survey reveal the following marketer
views on MTA:
Low Satisfaction.
Marketers not happy with the data they are getting.*
Fragmented Provider Ecosystem.
Ten of 30 providers account for two-thirds of use.*
Mistrust and Hesitation.
Marketers have a huge lack of trust, and hesitate to use MTA.**
Minimal Expert Understanding.
Marketers don’t have MTA expertise.
Different Approaches.
Providers use different analytics methods and data, presenting
evidence in different ways.***
Sources:
* Quant survey of 118 marketers conducted by the MMA as part of this project
** Based on 15 in-depth one-on-one interviews with members
*** RFI response analysis
-29%
-35%
-30%
-25%
-20%
-15%
-10%
-5%
0%
MTA providers have a dismal Net
Promoter Score
Overall how likely is it you would recommend your main multi touch
attribution (MTA) provider? N=118, Total MTA users
1NPS style calculation created from 10 pt satisfaction question
23
24. And for Good Reasons, we found 25 different
modeling techniques - a prescription for chaos!
Confidential: Cannot be shared without permission from the Mobile Marketing Association
1. Agent-based modeling
2. Bayesian machine learning
3. Bayesian shrinkage
4. Control theory
5. Counter-factuals
6. Doubly robust propensity modeling
7. Elastic net
8. Ensemble based probabilistic
9. Experimental design
10. Frequent pattern analysis
11. GLM
12. Hidden stage Markov models
13. Hierarchical regression
14. LASSO
15. Last touch
16. Logistic regression
17. Monte Carlo simulation
18. Probability of exposure
19. Shapley values
20. Structural equation models
21. Survey based measurement
22. Time decay
23. Time series
24. Utility theory
25. Vector autoregression
24
25. But What is the
MMA Doing About It
Everything We Can to
Drive Successful
Adoption by Marketers
25
Greg Stuart
greg@mmaglobal.com
+1 631 702 0682
26. In September 2016, we Launched…
26
MATT is a community of industry experts committed to rethinking
the world of marketing measurement and attribution; seeking to
give marketers better measurements, tools and confidence in
connecting marketing to business outcomes.
All MMA members are invited to participate in MATT
27. Applied to
34.7% of
Campaigns
(on avg.)
34%
Already
Have an
MTA
Solution
MATT’s MTA Mission…
This is where we started
-29
NPS!
27 Based on MMA Survey from November 2016.
28. MATT & MTA
Governance Global and NA Board MTA Marketers Council
(24+ members of MMA Boards)
MMA Board
MTA Steering Executive Committee
MMA Attribution
Marketer Advisory Task Force
MMA Attribution
Technical Advisory Task Force
Role:
* Vet and validate methodology/math
* Provide input on methodologies that are trustworthy and powerful,
validation evidence that is believable, data quality that makes
providers’ systems dependable and accurate
1-800-Flowers.com, Allstate, American Express OPEN, Bank of America,
Chobani, Choice Hotels, Colgate-Palmolive, Dunkin Brands, E*TRADE, Ford
Motor Co, General Motors Corp, GlaxoSmithKline, Johnson & Johnson, JP
Morgan Chase Bank, Marriott International, Mastercard Worldwide,
MillerCoors, Target Brands, Inc., The Wendy's Company, T-Mobile, Uber,
Unilever
1-800-Flowers.com, Allstate, American Express OPEN, Bank of America,
Chobani, Choice Hotels, Colgate-Palmolive, Dunkin Brands, E*TRADE,
Ford Motor Co, General Motors Corp, GlaxoSmithKline, Johnson &
Johnson, JP Morgan Chase Bank, Marriott International, Mastercard
Worldwide, MillerCoors, Target Brands, Inc., The Wendy's Company, T-
Mobile, Uber, Unilever
MMA Staff
MTA Subject Matter Expert (Joel Rubinson)
MATT Director (Mike Lewandowski)
MMA Global Board
of Directors
MMA North
America Board of
Directors
MATT Working Groups
1 Data Quality &
Accuracy
2 Mapping Data
Readiness
3 Addressing
Walled Gardens
Tiger Teams
New
Phase III
Role:
* Advise on the deliverable to MMA membership and industry at large
* Answers the question “What do I need from MTA that will impact
marketing decisions on digital, mobile and marketing spend?”
28
29. The MTA Steering Executive Committee – comprised
of Global & NA Board Members -- is guiding the initiative
Luis Di Como
SVP/Global Media
Unilever
Lou Paskalis
SVP/Enterprise Media
Planning, Investment &
Measurement Exec.
Bank of America
Amit Shah
CMO
1-800-Flowers.com
Sanjay Gupta
EVP/Marketing, Innovation
& Corporate Relations
Allstate Insurance
Company
29
30. GlobalBoardKicksOffProject MATT’s Major Process Milestones, so far…
CMO
Review
of RFI
submissions and
1:1 interviews
Provider Evaluation MTA Selection Tool Development
RFI to
Providers
500 pages of
Responses
from 19
qualified
providers
15
In-depth
Discussions
+ Academics
Quantitative
Survey I of
Marketers
N = 190
Input from
Joint
MTA Board
Committees
Guidance for
Marketers on how
to choose the right
providers
Marketer Needs Assessment
MTA Board
Task Force
Feedback
Quantitative
Survey II of
Marketers
N = 400
Technical
Analytics
Experts
Analytics
Expert
Land-
scape
Report
How-To
RFI
MTA
Scoring
Tool
Glossary
30 Hours
Phone
Review
4-Part MTA
Webinar Series
for Marketers
2016 2nd Half 2017 1st Half
MTA
Provider
Interviews
Working Group Formation
30
Data Quality
& Accuracy
Working
Group
MTA
Success
Workbook
2017 2nd Half
Color Key: Light Blue = Inputs from Marketers and Vendors
Aqua = Inputs from Survey, RFI, or Interviews
Dark Blue = Outputs in form of MTA Acceleration Tools
Walled
Gardens
Working
Group
Input
from 78
Member
Marketers
Data
Strategy
Guide
Data
Acquisiti
on RFI
Marketer
Survey III
MATT Data
Map 2.0
WG
Position
Paper
31. Started with “Selecting…” Launched in 2016
31
* The Decision Guide is available for members only at http://www.mmaglobal.com/matt.
The Report:
A comprehensive
guide to MTA
MTA RFI Template
Scoring Tool:
To help with
evaluation
4-Part MATT MTA
Webinar Series
Part 1:
Intro to Multi-Touch Attribution
(MTA) Methods
Part 2:
Selecting the Best MTA
Provider For Your Needs
Part 3:
Making Sense of Attribution
Approaches
Part 4:
Leveraging MTA to Improve
Marketing Effectiveness
32. Category
Elements of
Attribution Solutions
Education
Definitions
Needed
Codeof
Conduct
Best
Practices
Validation
Research
Standards
Auditing
Parameters
Data Quality
Data Quality & Accuracy (and Walled Gardens) X X X X X X
Standardize and Leverage Unified IDs X X X X X
Single Source Linkage to Sales Data X X X X
Analytic
Validity
Experimental Design X X X X
Transparency of Approach X X
Validation of Results and Outcomes X X X
Business
Outcomes
Facilitate Agile Marketing X X
Enhance Mobile Readiness X X X X X X X
Prove Lift in Campaign Performance X X X
Solution
Completeness
Specific Approach for Offline Media X X X X X
Use MTA for Both Brand and Performance Goals X X X
Comprehensive Answers for Planning and Budgeting X X X
32
MTA Has 12 Elements that Need Solutions
33. Ultimately, the information we gathered from the process
produced six main Use Cases for MTA
Confidential: Cannot be shared without permission from the Mobile Marketing Association
Budgeting and Planning
Agile (optimization of)
Marketing
Offline
Business
Online
Business
Online and
Offline Business
Offline
Business
Online
Business
Online and
Offline
Business
33
In the provider assessments, we have grouped vendor
leaders by the following use cases:
34. We then evaluated the importance of the 7 scoring
modules across each use case
Confidential: Cannot be shared without permission from the Mobile Marketing Association34
Offline Business Online Business Omni Channel
Seven Scoring Modules Planning
Marketing
activity
optimize
Planning
Marketing
activity
optimize
Planning
Marketing activity
optimize
1. Agile Marketing - 30% - 30% - 35%
2. Mobile ready 10% 15% 10% 15% 10% 15%
3. Comprehensive across marketing
channels, goals, process stage
30% 5% 20% - 30% 10%
4. Appropriate for offline businesses 20% 15% - - 15% 5%
5. Appropriate for online businesses - - 30% 20% 15% 5%
6. Offline media effects 10% 5% 10% 5% 10% 10%
7. Validity 30% 30% 30% 30% 20% 20%
35. Moved into “Applying…” Launching 2017
35
MTA Success WorkbookMTA Data Strategy Guide Data Acquisition RFI Marketer’s Position Paper
on Walled Gardens
MATT Data Map
36. Coming Sept: Data Map (Visual)
36
Applying
The Data Map helps to overview the strategy
conversations
• Gives framework for more complete picture
• Value is collaborative tool for marketer,
agency, DMP, and MTA provider to use to
ensure nothing gets left out of the
discussion
Think about major buckets of data more
clearly
1. Linkable
2. Aggregate
3. Profiling
4. Conversions
42. Again, the Short Version of What is MTA…
42
Multi-Touch Attribution: The science of
using advanced analytics,
• on user level data,
• to allocate proportional credit,
• across a granular list of marketing
touchpoints across many,
• and hopefully all, online and offline channels,
leading to a desired customer outcome.
Excluded: Traditional MMM, brand tracking and last-touch attribution methods
43. So, let’s break it down into a What, How,
Why review of MTA, as follows:
1a: WHAT Multi-Touch Attribution IS
1b: WHAT Multi-Touch Attribution IS NOT
2a: HOW does it work
2b: HOW do you get there
2c: HOW difficult is it to do
2d: HOW much does it cost
3: WHY do it
44
44. What, How & Why of MTA
45
1a: WHAT Multi-Touch Attribution IS
I. Documented Definition: Science of using advanced analytics on user level data to
allocate proportional credit across a granular list of marketing touch points across
many and hopefully all online and off-line channels, leading to a desired customer
outcome.
II. Or put another way: MTA is an analytical method to distinguish what's working when
dozens of marketing activities and approaches are occurring at the same time in order
to estimate what is driving incremental lift.
i. Or, to measure impact so as to do more of what is working and less of what is not,
therefore, increasing marketing productivity.
CONFIDENTIAL
45. 46
1b: WHAT Multi-Touch Attribution IS NOT
I. Does not include methods that exclusively use Media Mix Modeling (MMM).
I. Also, is not brand/campaign tracking, or pop-up type surveys a result of ad exposure.
II. Is NOT limited to measurement of only whole media channels (TV, Radio, Digital, etc.)
in aggregate as sufficient.
III. It is NOT last touch measurement.
IV. Is NOT retrospective, requiring 2-3 years of back data to make decisions on today’s
media allocations.
V. Is NOT restricted to paid advertising as it can include the effect of owned media page
views and even social media under certain circumstances.
What, How & Why of MTA
CONFIDENTIAL
46. What, How & Why of MTA
47
2a: HOW does it work
I. Calculates the probability that a user will convert (brand, sales, other) based on
exposure to a combination of marketing activities based on modeling the behavior of
individual users.
i. Requires a large data set, often with millions of records, linking marketing activities and
outcomes for the same user.
II. Traditional media variables can be put into these user level models based on
probability of exposure...not optimal.
i. Often the actual exposure event cannot be observed (e.g. if a given user saw a
commercial on a specific TV show.)
III. Analyzes all events that potentially affected the conversion from the start of the
shopper journey.
CONFIDENTIAL
47. What, How & Why of MTA
48
2b: HOW do you get there
I. While some ‘tech-centric’ marketers have built their own solutions, most use one of
~20 MTA providers that might work with the media agency and DMP to construct the
necessary data connections.
i. MMA examined 19 providers using over 2-dozen, fundamentally different statistical
approaches, with limited or no validation.
II. All MTA projects are first and foremost ‘Big Data’ projects *
i. Must create critical data linkages and elements that many companies might not yet have
in place (tagging, profiles, server logs, offline sales, etc.).
CONFIDENTIAL
* See visual of MTA Data Map
48. What, How & Why of MTA
49
2c: HOW difficult is it to do
I. Hard. Largely because marketers don’t have a data strategy in place and there is a lot
of complexity to setting that up.
i. There is a need to test all the connections and there are a lot of issues with getting good
data, especially in light of the walled gardens, sources of sales data and more.
ii. It is often a 6+ month process just to set up the data and fully test all the data linkages.
II. MTA also suffers from a lack of trust when it produces results contrary to MMM
because the ways of determining the validity of MTA output are evolving.
i. Is the company ready to sacrifice “sacred cows”.
CONFIDENTIAL
49. What, How & Why of MTA
50
2d: HOW much does it cost
I. Costs vary, but based on a
recent MMA survey, full MTA
service providers will probably
charge a minimum of $250,000
for a proof of concept study
II. Costs can run to low single digit
millions annually for a single
business in a single market (e.g.
retailer marque, in the U.S.)
6 our of 10 marketers
expect to pay up to $500K *
13.60%
52.50%
8.50%
8.50%
3.40%
8.50%
5.10%
<$100k
100k to 500k
500k to 1 million
1 to 2 million
Over 2 million
I prefer not to answer
Don´t Know
* MTA Survey #3, March 2017: “What do you estimate the out of pocket costs were for your organization’s MTA
services in 2016? If you didn’t use MTA in 2016 but plan to use it in the future, what do you estimate the out of
pocket costs for your organization’s MTA services will be for a full year?”
CONFIDENTIAL
50. What, How & Why of MTA
51
3: WHY do it
I. Built to guide marketers to hit the bulls eye of right consumer, right time, right message, right
screen, to improve marketing return.
i. SMoX fundamentally proves that MTA is critical to the future of marketing productivity.
II. There is no other approach built from the ground up to reflect what makes marketing
different in a data driven digital age. And it appears to work.
III. Estimates optimal allocations at both the media channel level and at more granular and
tactical levels.
IV. Some suggest that the value is in the range of a 15-25%+ increase in marketing productivity
(sales lift during campaigns for example).
V. MTA encourages better marketing!. It gives credit to up-steam/top of funnel activities that
deserve credit. Without this, e.g. by using last touch, marketers would be falsely guided to
purely promotional and tactical messages, mortgaging a brand's equity and its future.
CONFIDENTIAL
51. Critical Focus Areas Ranked by Marketers
52
2%
3%
7%
5%
5%
7%
8%
5%
16%
15%
28%
5%
15%
7%
5%
7%
10%
5%
5%
16%
13%
13%
3%
5%
11%
18%
10%
7%
8%
13%
3%
8%
13%
2%
7%
3%
8%
11%
13%
18%
13%
8%
10%
7%
10%
3%
11%
5%
11%
8%
7%
13%
8%
11%
11%
0% 10% 20% 30% 40% 50% 60% 70% 80%
Enhance Mobile Readiness
Breaking down “Walled Gardens”
Data Quality and Accuracy
Standardize and leverage Unified IDs
Single source Linkage to sales data
Facilitate Agile Marketing
Specific approach for Offline Media
Transparency of Approach
Validation of results and outcomes
Use MTA for both Brand and Performance goals
Prove lift in Campaign Performance
Most Important 2nd 3rd 4th 5th
Below is a list of areas that marketers have identified as important priorities to
improve the value they receive from MTA solutions. Which of these priorities
would make MTA more valuable to you and your organization. Please rank the
top 5 with 1 being most important.
Marketers
need
CONFIDENCE
52. Started with “Selecting…” Back in Nov 2016
53
Selecting
* The Decision Guide is available for members only at http://www.mmaglobal.com/matt.
The Report:
A comprehensive
guide to MTA
MTA RFI Template
Scoring Tool:
To help with
evaluation
4-Part MATT MTA
Webinar Series
Part 1:
Intro to Multi-Touch Attribution
(MTA) Methods
Part 2:
Selecting the Best MTA
Provider For Your Needs
Part 3:
Making Sense of Attribution
Approaches
Part 4:
Leveraging MTA to Improve
Marketing Effectiveness
Background:Output
53. Original 12 Elements of “Good MTA”
54
Elements of
Attribution Solutions
Education
Definitions
Needed
Code of
Conduct
Best
Practices
Validation
Research
Standards
Auditing
Para-
meters
Transparency of Approach X X
Validation of Results and Outcomes X X X
Standardize and Leverage Unified IDs X X X X X
Prove Lift in Campaign Performance X X X
Specific Approach for Offline Media X X X X X
Facilitate Agile Marketing X X
Single Source Linkage to Sales Data X X X X
Comprehensive Answers for Planning & Budgeting X X X
Enhance Mobile Readiness X X X X X X X
Use MTA for Both Brand and Performance Goals X X X
Data Quality & Accuracy (and Walled Gardens) X X X X X X
Experimental Design X X X X
Background:Insight
54. Data Strategy Guide
55
MTA Data Strategy Guide provides:
• Provides a section by section explanation of what should
be considered in the MTA data planning process
• Defines Data Asset Types and Relational Structures
• Defines Linking of different types of data
Value to Marketers
Value is most researchers, marketers, and analytics people
are not data architects so this guide ensures that the team
understands the data assets it will take to be successful with
MTA.
Applying
“MTA is first and foremost a big data project and many big data projects fail…” - MTA Provider
2 of 5
55. MTA Success Workbook
56
MTA Success Workbook provides:
• Contains decision trees to guide marketers
through the data strategy elements
• Offers an excel workbook for capturing the
details of their company strategy
• When complete, the is the “Roadmap”
needed for successful deployment
Value to Marketers
Value is a workbook that allows the marketer to
take the principles of a data strategy for MTA and
create the actual plan for their company
Applying
“MTA is first and foremost a big data project and many big data projects fail…” - MTA Provider
ID structure of your DMP strategy for unified IDs Analytic approach
cookie/unified/other
use DMP, use MTA provider,
other Unified ID platform don't
plan to have this
will use experimental
design rather than MTA
3 of 5
56. MATT MTA: Board Discussion
57
1. Last board meeting we “level set” the What,
How, and Why of MTA (it’s in the pre-read).
2. What can you can you share about MTA efforts in
your organization?
3. How can MMA be most effective in our path to
push MTA adoption and increase confidence?
1. What has MMA done that has been helpful?
2. What else can MMA do at the industry level?
4. What more can we do to help you & your team?
5. Will you support the MTA “Acceleration Plan” to
be shared today? Financially?
Today’s Agenda:
1. Quick review of
program &
objectives.
2. Program update on
key successes.
3. Vision for what
could come next.
4. Discussion of the
future of MTA and
what it means to
you.
57. Data RFI Template(s)
58
MTA Data RFI Template provides:
• Industry agreed to structure for major
questions to vendors
Value to Marketers
Value is offering a structure that will allow
clients to fast track their search for
targetable segments that will be valid and
useful
Applying
“Until we trust the data no one will accept the MTA analysis that is built on it” - Marketer
4 of 5
58. Marketer’s Rights on Walled Gardens
59
Marketer’s View on Walled Gardens provides:
• Identifies the shared challenges caused by
walled gardens
• Offers suggestions for working with and
around walled gardens
• Sets expectations for what is and is not
possible within walled garden environments
Value to Marketers
Will help marketers negotiate more effectively to
get the data they need and deserve access to n
order to have complete results from their MTA
study and to believe the results that come out
Applying
“We could only capture 11% of ad impressions on our first MTA study. #fail!” - Marketer
5 of 5
60. Mobile should be a third of the spend.
Optimizing within mobile has a huge upside potential
61
Upside
compared to
no mobile in
the mix
Upside from
also
optimizing
within mobile
124%
166%
OPTIMAL ALLOCATION OF
MOBILE IN THE MIX
UPSIDE FROM MOBILE IN
TOTAL CAMPAIGN
PERFORMANCE
12%
21%
38%
29%
Mobile Social
Other mobile
TV
Other
61. What is MTA, How
does it Work &
Why Do it
To be Shared at Board
Meeting
62
62. SMARTIES GLOBAL AND NORTH AMERICA
63
Business Impact Index – in
partnership with WARC
“Top 5” ranking of mobile
marketing companies –
including an individual ranking
each of agencies and brands –
that are achieving the biggest
business impact.
Submissions have been negatively
impacted by Publicis
announcement and WPP hacking
Deadlines have been extended
to help WPP companies submit
63. Review of MMA Purpose
WHO (The People We Serve):
Prime Audience: Chief Marketers
WHY (Our Reason for Being):
Purpose: To accelerate the transformation and innovation of marketing through mobile,
driving business growth with closer and stronger consumer engagement.
WHAT (Our Strategic Priorities):
Primary Focus:
1. Demonstrating Measurement and Impact: proving effectiveness and optimizing impact
2. Cultivating Inspiration: aimed at the Chief Marketer; guiding best practices and
driving innovation
3. Building Capability for Success: fostering know-how and confidence within the
Chief Marketer’s organization
Secondary Focus:
Advocacy – monitoring and maintenance activity only; via partnership with the DAA
64. Review of MMA’s Focus to Building New Marketing Channel
65
USP &
Positioning
Economic
Value & ROI
Getting it
Right/Best
Creative
Exploratory &
Effectiveness
Building
Measure-
ment Tools
Organiza-
tion Design
Coming in
2017
Wave II
Wave I
65. What Do we
Know from
Members
Recap of what we
learned from survey’s
and our benchmarks
66
66. • Selecting and on boarding an MTA provider is a long process, over 1 year from
initial RFP to MTA providers
• Non Users of MTA have higher estimates in terms of the selection time
needed (39% in 6-12 mo.) and on boarding time needed (49% in 3-6 mo.)
• Non users also have a lot higher expectation about the impact of MTA on
marketing ROI (36% anticipate above 10-40% ROI on average)
• Marketers pay/expect to pay up to $500K a year for MTA, but two thirds are
not sure if MTA ”pays for itself”
• Proving lift is the most critical priority for marketers
Summary: MTA Survey’s – Key Findings
67
67. State of Application: Of 34% of marketers that have MTA, they use it
selectively and many are unsure the benefits outweigh its cost.
35%
18%16%
17%
15%
Less than 10%
10% to 30%
30% to 50%
50% to 70%
70% to 100%
What share of your total marketing budget (including
advertising) do you assess roughly speaking, using a
Multi-Touch Attribution solution? (Please consider
whether you use MTA for all or just some of your
marketing activities and channels)
Yet, the majority are not convinced the
cost of MTA is worth the benefits
2%
10%
27%
43%
18%
Definitely not
Probably not
Not sure
Yes, probably
Yes, definitely
Does your current MTA solution pay
off for its cost, in terms of driving
incremental ROI and impact for
your marketing spend to justify its
cost? N=107,
Six out of 10 marketers think their
MTA solution, on average, drives
some incremental results or ROI
5%
28%
22%
11%
2%
32%
No increase
Up to 10% on average
Up to 10-20% on
average
Up to 20-40% on
average
40% and above
Not sure / Don’t Know
Based on your experience, what
do you think is the average
impact/lift of using your MTA
solution on the total ROI of your
marketing activities?
68
69% use on less than 50%
Majority of marketers use MTA selectively,
not across all of their budget
68. Cost vs. Benefits: Marketers pay/expect to pay up to
$500K a year for MTA, but 2/3rds are not sure if MTA
”pays for itself”
69
13.60%
52.50%
8.50%
8.50%
3.40%
8.50%
5.10%
<$100k
100k to 500k
500k to 1 million
1 to 2 million
Over 2 million
I prefer not to answer
Don´t Know
What do you estimate the out of pocket costs were for your organization’s MTA services in
2016? If you didn’t use MTA in 2016 but plan to use it in the future, what do you estimate the
out of pocket costs for your organization’s MTA services will be for a full year?
25%
10%
32%
32%
Fully pay(s) for itself and
return(s) a profit from
increased marketing…
Fully pay(s) for itself and just
break(s) even
Partly pay(s) for itself
Not sure
Would you say the MTA solution you currently use, pays for itself? If you
don’t currently use MTA, do you expect the MTA solution you will
implement, will pay for itself by end of Year 1?
6 our of 10 marketers expect to
pay up to $500K
Two thirds of marketers are not sure
if MTA pays for itself
69. Differences in Experience: Users & Non say it takes 6
mos. to pick a partner but non users say takes longer to
on board
70
How long did it take you /do you think it will take you to to
select your current MTA solution (from the date you started the
RFP process until the date you had a signed contract)?
3%
20%
38%
25%
15%
0%
0%
18%
33%
39%
0%
9%
0% 10% 20% 30% 40% 50%
One month
More than a month…
3-6 months
6-12 months
1 year or more
Dont Know
13%
13%
40%
23%
13%
0%
9%
15%
49%
18%
3%
6%
0% 20% 40% 60%
One month
More than a…
3-6 months
6-12 months
1 year or more
Dont Know
Users
Non Users
How long did it take you /do you think it will take you to to
onboard your current MTA solution (from the date you signed
the contract)?
At least 6 months to select At least 6 months to onboard
70. Non users have a lot higher expectation about the impact of MTA on
marketing ROI
71
15%
33%
8%
5%
5%
35%
3%
27%
24%
12%
0%
33%
0% 10% 20% 30% 40%
No increase
Up to 10% on average
Up to 10-20% on average
Up to 20-40% on average
40% and above
Not sure / Don’t Know
Users
Non users
For the part of the budget where you will use MTA, what do you expect will be the impact on the
marketing ROI as a result of using MTA?
Hinweis der Redaktion
Behavioral: first, second and third party data sets where a person has been identified as in market auto shopper or going through another lifestage (married, kids, house) that indicates buying insurance.
Contextual: targeting based on the type of content that is being displayed on the screen (car shopping, home shopping, lifestages etc.)
Both of these targeting methods delivered results for consideration. As we will see later, contextual did’t quite work for sales.
The proliferation of integrated individual-level data (i.e., targeting, behavior, other) transforms marketing and makes it possible to connect 1-to-1 with consumers.
This creates big performance gains in granular marketing tactics
Marketing Mix Modeling, which EVERYONE uses, only goes so far (just to the channel level)
It makes sense then that we use the same individual data for measurement that we use for targeting. That is what MTA is built to do.
Rinse & repeat.