Upload
emarketer
View
6.181
Download
1
Embed Size (px)
Citation preview
© 2015 eMarketer Inc.
Made possible by
Demystifying Mobile Attribution
Cathy Boyle
Senior Analyst, Mobile
October 8, 2015
© 2015 eMarketer Inc.
Agenda
Defining attribution
The state of mobile attribution
Methods and models applied
Five missing links with mobile
Two unique elements to consider
The value of mobile (three brand examples)
© 2015 eMarketer Inc.
A Working Definition of Attribution:
A method of assigning credit to a particular marketing
interaction, brand touchpoint or channel in order to
quantify the contribution that interaction, touchpoint or
channel makes toward a desired business goal
© 2015 eMarketer Inc.
The State of Mobile
Attribution
© 2015 eMarketer Inc.
Marketing is as fractured across devices as
consumers’ attention
Data from Google’s servers indicates, “90% of people use multiple screens
sequentially to accomplish a task.”
Source: July 2015 interview with Surojit Chatterjee, Product Management Director, Mobile Search Ads and AdSense for Search, Google.
© 2015 eMarketer Inc.
Marketers are unsure how to assign credit to
mobile marketing and media touchpoints
“The way people transition between
devices and mediums really puts a
strain [on resources]. Our analysis
shows consumers have to see
something four times before they take
action on it. How do we value
each one of those
engagements? It's increasingly
difficult.”
—Gabriella Weinstein, senior mobile product
marketing manager at Rue La La
© 2015 eMarketer Inc.
As a result, attribution has become downright
difficult
Many marketers
are not
measuring the
performance of
mobile, let alone
tackling
attribution
Roughly half
of those
polled were
NOT
measuring
mobile
engagement
or return on
investment
© 2015 eMarketer Inc.
The limited ability to track mobile ROI is holding
marketers back from increasing mobile budgets
© 2015 eMarketer Inc.
But it’s not impossible to measure and assign
credit to mobile, along with other channels
“I get that it’s hard. But it’s whining.
Competitive advantage never came
from having perfect information. The
opportunity for brands to get a
leg up on their competitors is
now [before models and systems are
perfected].”
—Greg Stuart, CEO of the Mobile Marketing
Association (MMA)
© 2015 eMarketer Inc.
Attribution Methods Used
When Mobile Is Part of the
Marketing Mix
© 2015 eMarketer Inc.
Mobile’s contribution can’t be measured in a
vacuum
“Doing any kind of attribution in a silo
is a huge mistake because you miss a
large part of what is driving user behavior.
You have to look at attribution
holistically in order to get a good view of
the value of all your media.”
—Phil Gross, vice president of product
management at Visual IQ, an attribution provider
© 2015 eMarketer Inc.
Marketers are blending two attribution methods
To quantify the contribution made by mobile, along with other channels
and devices, marketers typically employ two approaches:
1. Top-Down Modeling (marketing/media-mix modeling)
– Input media spend for all channels, formats, gross rating points (GRPs), print exposures and
other marketing impressions into a model
– Can add in additional data points, including sales or external influences
2. Bottom-Up Modeling (path analysis)
– Typically used to attribute digital advertising channels and formats, often down to the impression
or ad creative level
– Used as a means to optimize campaigns in as close to real time as possible
– Used as a means of justifying and planning future buys
© 2015 eMarketer Inc.
Two “crediting” methods are used for path
analysis—one simplistic, the other complex
1. Last-click attribution: An approach that assigns all conversion
credit to the last marketing touchpoint a consumer interacted with
prior to taking a desired action
2. Multitouch attribution: An approach that allocates credit
proportionally to marketing touchpoints a consumer interacted with
across platforms, channels and devices prior to the desired action
© 2015 eMarketer Inc.
Last-touch attribution is the most prevalent
method used for assigning credit to mobile
“We value and see the market moving
toward a multitouch attribution
model. The reality is the industry is a ways
out from being able to support that from an
infrastructure and a billing perspective.
… It’s going to be a while before the
industry is really able to handle that model
on a broad basis.”
—Justin Landis, director of content marketing for
mobile attribution provider Kochava
© 2015 eMarketer Inc.
Five Missing Links and
Two Unique Elements
© 2015 eMarketer Inc.
Mobile attribution difficulties stem from the
absence of five key elements
1. A single identifier to serve as a proxy for a person
2. Broad-scale access to impression-level data
3. Large-scale campaigns and large budgets, which produce
large data sets for performance analysis
4. Clarity on the right metrics to measure
5. Websites as the primary destination and point of sale
© 2015 eMarketer Inc.
In addition, two unique elements in mobile need
to be considered when it comes to attribution
1. The app stores as gatekeepers
2. Phone calls as an integral means of response
Examples bring the challenges and solutions to life…
© 2015 eMarketer Inc.
Example Goal: Raise Brand Awareness
Target audience: Mobile mothers ages 24 to 35
Media Mix: Mobile display ads served in-app and
on mobile websites
Devices: Smartphones and tablets
Three key performance indicators (KPIs):
Visits to the Love-Me-Feed-Me mobile website
Visits to the Love-Me-Feed-Me app
Reach and frequency of brand exposures among
that target audience
Company:
Pet food manufacturer
(fictitious)
Image source: Clipart Panda
Love Feed
© 2015 eMarketer Inc.
Challenge #1:
Conducting path
analysis without the
comfort of cookies
In-App ID Web ID
Mobile has unique app and web ID’s
To report accurate reach, frequency of
exposure and clickthroughs, app and mobile
web ID’s need to be “stitched”
together
Exception: Publishers with user-generated
ID’s that span channels
User ID
Others
© 2015 eMarketer Inc.
Challenge #2:
Mothers likely
use both a
smartphone
and tablet
58% of US
smartphone users
also use a tablet,
according to eMarketer
estimates Smartphone
ID’s
Tablets
ID’s
© 2015 eMarketer Inc.
Challenge #3:
More limited
sharing of
impression-level
data in mobile
vs. desktop
Number of viewable impressions
served prior to the last click
…when they were served
…to which sites or apps
…to which device IDs
“Brands care about different
metrics, not just click. They
want to understand the
effects of exposure
data, and there are
challenges today around
measuring that in mobile.”
—Rob Friedlander, co-founder
and CMO of Phluant Mobile, an
advertising solutions provider
Not always
provided
© 2015 eMarketer Inc.
Mobile differs from desktop in the number of
firms with Media Ratings Council accreditation
“Yes, mobile impression data is hard.
The proof is that only one company has
been MRC accredited to deliver and
count [display, rich media and
video impressions and clicks]
correctly in mobile over the last two
years.”
—Richy Glassberg, CEO of Medialets,
an MRC-accredited buy-side ad server for mobile
display, rich media and video impressions
© 2015 eMarketer Inc.
Example Goal: Drive App Downloads
Company:
Boutique Hotel Chain
(fictitious)
Image source: Clipart Panda
CuteSpot
Target audience: Recent first-time guests
Media Mix: Paid and owned mobile media
and marketing
Two KPIs:
The number of installs generated by the target
audience
The number of bookings made from the new
app users
© 2015 eMarketer Inc.
Challenge #1:
The app store
being in the
middle of the
conversion
process
To properly credit the
source of the install:
The ID detected from the
ad is, in most cases,
matched against the ID
detected when the app is
first opened
CuteSpot Install Now
Get Two
Free Nights
in Paradise
Suggested AppAd Tracking
via:
App ID
Web ID
Probabilistic
ID
Install Tracking
via:
App ID
Referrer
(Android)
No tracking within
the app stores
© 2015 eMarketer Inc.
Challenge #2:
Understanding
the influence
of the app’s
profile in the
app store
At this point, the affect
the app profile has on
conversion rates of
app-install ads is not
directly quantifiable
CuteSpot Boutique Hotel
Bequia, Grenadines
© 2015 eMarketer Inc.
Challenge #3: The scope of analysis is often
limited to purchases made only in the app
“Through reporting mechanisms
advertisers can connect
all the dots. So if the [app]
user goes on to make a
purchase on the website or in
the store, they can report that
back to us and have that
included in the data about that
user.”
To gauge the true value of
the install and the media
source, all bookings
made post-install need
to be taken into
consideration
—Daniel Kahtan, director of sales and
business development at AppsFlyer, a
mobile app attribution provider
© 2015 eMarketer Inc.
Example Goal: Increase sales
Company:
Home goods retailer
(fictitious)
Image source: Clipart Panda
Nest Nuggets
Target audience: Recent buyers
Media Mix: Paid and owned mobile media and
marketing
Two KPIs:
Increase mobile commerce sales from their mobile
storefronts (mobile website and app)
Increase in-store sales
© 2015 eMarketer Inc.
Challenge #1: The length of the attribution
window
“If they go down the app
download path, they may find
that their conversion rates
drop vs. just sending them to
the mobile web. They may want
to look at a longer-term horizon,
because that [app] user may
have a greater lifetime value.”
—Jason Hicks, vice president of
marketing and client services at
Kochava
“Best experience” “Acceptable experience”
Shorter
path to purchase
Longer
path to purchase
if app not already
installed
© 2015 eMarketer Inc.
Challenge #2: The impact phone calls have on
conversions needs to factored in
93 billion calls will
be made to
businesses in the US
from smartphones in
2015
Up 23% from 2014
© 2015 eMarketer Inc.
Challenge #3: How to quantify the impact of
mobile on in-store sales
Mobile devices influenced
or helped to convert about $970 billion in
US sales
Represents 28% of the
2014 total
“You would be shocked how
many marketers still don’t know
that you can trace an ad
on a mobile phone all the
way to an offline
transaction in a store.”
—Tim Jenkins, CEO of 4Info
© 2015 eMarketer Inc.
“In some of the cases, brands only had 1%
of their budget going into mobile.
That may drive behavior, but it becomes
almost a rounding error in some of
these models compared to what television
might be influencing.”
—Rex Briggs, founder and CEO of Marketing
Evolution
Challenge #4: Limited data sets produced by
mobile campaigns
© 2015 eMarketer Inc.
The Value of Mobile
for Four Brands
© 2015 eMarketer Inc.
Mobile was second only to TV in creating the image that
MasterCard was a “good card to carry when traveling”
Per dollar spent, mobile worked almost twice as hard
(1.7x) compared with the campaign average in respect to building
this brand image
Source: Mobile Marketing Association (MMA), “Smart Mobile Cross Marketing Effectiveness (SMoX) Study,” May 2015
© 2015 eMarketer Inc.
Mobile exceeded broadcast and cable TV in creating the
intent to shop
Mobile drove 14% of the change in overall shopping intent
despite [receiving] only 7% of the spend
Source: Mobile Marketing Association (MMA), “Smart Mobile Cross Marketing Effectiveness (SMoX) Study,” May 2015
© 2015 eMarketer Inc.
Mobile drove 25% of “top-of-mind” awareness
6% of sales with 5% of [the overall] budget
Compared with the campaign average, mobile generated nearly five
times (4.8x) the level of top-of-mind awareness
Source: Mobile Marketing Association (MMA), “Smart Mobile Cross Marketing Effectiveness (SMoX) Study,” May 2015
© 2015 eMarketer Inc.
Key takeaway: All the brands could increase
their ROI by shifting more money into mobile
“Given what we've seen so far
[from the SMoX study], most
brand categories should be
spending more in
mobile.”
—Rex Briggs, founder and CEO of Marketing
Evolution
© 2015 eMarketer Inc.
Conclusions
There is widespread agreement that attribution with mobile in the mix is a complex and resource-intensive undertaking
The pitfall to avoid is believing there is a perfect methodology,
tool or answer
Any move that will provide more clarity on mobile’s
contribution is worth taking, as evidenced by the results of the
brand studies
© 2014 Adobe Systems Incorporated. All Rights Reserved.
Andrew KoperwasProduct Marketing Manager
Twitter: @koperwas
© 2015 Adobe Systems Incorporated. All Rights Reserved.
Adobe Analytics – Mobile App Analytics
• App-centric user interface for mobile teams lets you use:
• Pathing analysis to understand which screens and UI elements are more frequently used
• Cohort analysis to understand behavior of cohorts over time who download app during the same period
• Analyze user acquisition by:
• Campaign source
• Specific campaign
• Source and device model
© 2015 Adobe Systems Incorporated. All Rights Reserved.
Adobe Analytics Cross-Channel Attribution
• Stitch together marketing and
customer data from both offline and
online sources
• Ability to use rules-based models
(e.g. first/last, u-shaped) or algorithmic
attribution
• Analyze customer interactions (not
just campaigns) – across paid, earned
and owned sources
• Combine with all of Adobe Analytics –
• Anomaly detection
• Contribution analysis
• Segmentation
• Propensity scoring
© 2015 eMarketer Inc.
Learn more about digital marketing with an
eMarketer corporate subscription
Around 200 eMarketer reports are published
each year. Here are some recent reports and
an upcoming webinar you may be interested in:
Q&A Session
Made possible by
You will receive an email tomorrow with a link to view the
deck and webinar recording.
To learn more: www.emarketer.com/products
800-405-0844 or [email protected]
Cathy Boyle
Demystifying Mobile Attribution
Cross-Platform Attribution 2015: Device Identification, Big Data
Pose Continued Challenges
Cross-Device Targeting: Success Hinges on Device
Identification Methods
Mobile Audience Targeting: Have Industry Advances Raised
Advertisers' Confidence Levels?
Upcoming Webinar: Making Attribution Work in the
Age of Big Data, October 22, 2015