Big Data and Mobile Analytics - MMA SF Jan13

Mobile Marketing Association

LEVERAGING BIG DATA IN MOBILE MARKETING

Jennifer VeesenmeyerMerkle Analytics


• Merkle’s mobile division

• Award-winning mobile web, app, tablet and SMS solutions

• Innovative Responsive Design services

• VP in Merkle Analytics, Digital Analytics Practice(I own Mobile, Social and Site)

• 175 statisticians and analysts, >95% with a master’s degree +

Merkle, Mobile & Me

2


What we’ll cover today

Deadly Sins of Mobile

Analytics

Table Stakes for

Mobile Analytics

Big Data Ideas for Mobile

Marketing

Context & Level-Set Application

3




Analytics

Table Stakes for

Mobile Analytics


Marketing

4


7 Deadly Sins of Mobile Analytics

1. Lumping smartphones and tablets together

2. Measuring mobile marketing in a silo

3. Ignoring the context of purchase process

4. Measuring usage rather than impact

5. Reporting on mobile as a channel (when you’re using it as a platform)

6. Guessing instead of testing

7. Settling for subpar analytics

5


DEADLY SIN # 1 :

Lumping smartphones and tablets together

Research confirms what many of us have seen in our own data –smartphones and tablets are used differently.

Smartphones are an

on-the-go device, but

tablets are more for

media consumption.

A 2012 study by Flurry

Analytics also showed that

67% of time spent on

tablets was on games, but

only 39% on smartphones.

6


DEADLY SIN # 2 :

Measuring mobile marketing in a silo

Mobile ads increased

website conversions

by 36%.

In another study, Ghose

et al found that 33% of

mobile site visitors

made a store purchase

within 1 week.

Research has proven what we have suspected –mobile marketing improves conversions in other channels.

Impact of Cross-Media (mobile and web) Ads (Ghose, Han and Park 2012)

7


DEADLY SIN # 3 :

Ignoring the context of purchase process

Consumers take a multi-device path to purchase…and they use mobile devices throughout the consumer lifecycle.

Google Multiscreen Study (August 2012)

8


DEADLY SIN # 4 :

Measuring usage rather than impact

Getting to ROI is difficult

and many companies

don’t try.

This is an opportunity for

mobile marketers to gain a

competitive advantage.

Measuring usage is important for improving user experienceand understanding adoption, but it isn’t a measure of success.

Does your organization have a method for quantifying ROI

from mobile marketing programs ?

9


DEADLY SIN # 5 :

Reporting on mobile as a channel

When mobile has a

unique strategy, then it is

a channel to be reported

separately.

But, if it is an add-on to

your other marketing

initiatives, then it is a

platform to be compared.

If your mobile site and pc/laptop site have the same objectives, put them in the same report so you can compare performance.

(when you’re using it as a platform)

10


DEADLY SIN # 6 :

Guessing instead of testing

Optimize mobile

performance post

launch by starting

with the easiest first

– messages and

landing pages.

Testing is important both pre and post launch.

Did not perform any mobile testing or optimzation

Optimization without testing

Not applicable/don’t know

Data analysis of past campaign metrics

Customer feedback/survey

Testing (A/B split, multivariate)

What mobile optimization efforts were utilized by your organization in 2012?

11


DEADLY SIN # 7 :

Settling for subpar analytics

Have a plan (and budget) for

evolving your mobile analytics

capabilities.

Mobile marketing budgets

are increasing dramatically,

but they would increase even

faster if you had good analytics.

Limitations in mobile analytics should be acknowledged but not accepted.

12




Analytics

Table Stakes for

Mobile Analytics


Marketing

13


5 Table Stakes for Mobile Analytics

1. Tools: Get the right analytics tools in place

2. Metrics: Know what is important to measure

3. Process: Establish an end-to-end analytics process

4. Research: Understand your mobile customers

5. Plan to Integrate: Have a plan to integrate

mobile into your overall measurement strategy

These are the bare minimum capabilities for

mobile analytics. Seriously. If you don’t have

these in place, you’re behind.

14


Incumbent or niche partners

Appetite for learning curve and/or decentralized systems

Acceptance of nascent ecosystem

Sample Decision Types

Behavioral

Attitudinal

Competitive

Infrastructure Components

TOOLS:

Get the right analytics tools in place

Determine what the right analytic tools and capabilities are for your organization.

15


METRICS:

Know what is important to measure

Separate KPIs and diagnostic metrics from attributes. You need both for analysis, but not for reporting.

250+ attributes can be collected, depending on the analytics vendor chosen.

Only a handful are measures of success. Fewer still provide a valuable basis

for segmentation in dashboard-level reports.

BUT, getting your data at the user-level opens up many possibilities for

providing micro-targeted messaging. Thanks to Big Data technologies,

sometimes more data is more actionable.

It’s ironic that sometimes the cure for

drowning in data is more data.

16


PROCESS:

Establish an end-to-end analytics process

Mobile analytics starts upstream with insight and measurement strategy. Post-launch begins a continuous improvement process.

17


RESEARCH:

Understand your mobile customers

• Are your consumers engaging with content

differently now by each of the channels?

• Are your consumers changing how they

visit your web site?

– Review current trends over the past

few years. Is the trend accelerating?

• Are more consumers using search from

their mobile devices?

Internal research can help you understand the current need and forecast the opportunity.

18


PLAN TO INTEGRATE:

Have a plan to integrate mobile into your overall measurement strategy

An integrated measurement strategy includes:

• KPI framework with Global KPIs that apply

to all channels in addition to channel-

specific metrics

• Technology architecture for managing the

data and connecting multi-channel

behaviors to customers

• Reporting strategy that encourages cross

functional and multi-agency collaboration

You need an integrated measurement strategy –even if you don’t have an integrated marketing strategy yet.

19




Analytics

Table Stakes for

Mobile Analytics


Marketing

20


Image from WebOptimeez (http://bit.ly/VqJda9)

Examples of Big Data in mobile marketing:

• Multichannel CRM

• Location

• Social

21

Two keys to successfully using Big Data in mobile marketing:

• Get personal

• Go real-time


Big Data Ideas for Mobile Marketing

Multichannel CRM Data

1. Enriching customer profiles with mobile data

2. Targeting the omnichannel customer

3. Integrating mobile display, mobile web and email

Location Data

4. Testing coupon value vs. proximity

5. Predicting traffic patterns and using “behavioral fencing”

6. Researching cross-shopping, mobile-instore dynamics, motivators, etc

Social Data

7. Experimenting with Social MRI

22


MULTI-CHANNEL CRM DATA:

Enriching customer profiles with mobile

In ConnectedCRM we bring mobile data into the customer database to build toward an omni-channel view of the customer.

23

The analytics behind ConnectedCRM is explained in our upcoming book:


MULTI-CHANNEL CRM DATA:

Targeting the omnichannel customer

Data used:

• Product purchase history

• Average purchased product value

• Average monthly spend

Execution:

• Personalized, on a 1-1 level, SMS and In-App Coupons Available on a weekly basis. Delivered at strategic times of day and day of week.

Results:

• Mobile coupons had a 35% redemption rate (5-10% is typical)

Case Study Example: 5th Finger used customer purchase history to create personalized coupons for a major retailer.

24


MULTICHANNEL CRM DATA:

Integrating mobile display, mobile web and email

Case Study Example: Merkle used targeted mobile display to acquire new email opt-ins for Motorola Mobility.

25

Data used:• Device information• Mobile site behaviors• Email address

Execution:

• Targeted display ads on the individual device level

• Users were directed to different creative experiences (and value props) for different devices

• Mobile landing pages encouraged users to sign up for email newsletter, when they did, we entered them into the database with device information

• Email retargeting


LOCATION DATA:

Testing coupon value vs proximity

Ghose and Span found that you need to offer a 10% discount for every 1km between the user and retail store.

Research: Disentangling Coupon Value vs Distance Proximity Using Randomized Experiments (Ghose and Span, 2013)

26


LOCATION DATA:

Predicting traffic patterns and using“behavioral fencing”

Research is being done to

leverage predictive analytics

to model the probability a

customer will take different

actions based on historical

proximity and behavior.

Behavioral fencing is the new geo-fencing.

Image from MartketingEasy.net (http://bit.ly/MU9d95)

27


LOCATION DATA:

Researching cross-shopping, mobile-instore dynamics, motivators, etc

Example: Locately

• Customer Opt-in: In exchange for gifts cards, consumers opt-in to share location data by installing app or by enabling location sharing.

• Targeted Surveys: Consumers earn rewards by filling out surveys capturing their thoughts.

• Location Analytics & Processing: Places consumers drive past are tagged as missed opportunities.

New vendors are providing new ways to research.

28


SOCIAL DATA:

Experimenting with Social MRI

Behavio’s open-source Android platform turns phones into smart sensors of people’s behaviors and surroundings.

Behavio's Social MRI technlology was born at the MIT Media Lab, as part of research into understanding social and behavioral dynamics.

29


Let’s Review


Analytics

Table Stakes for

Mobile Analytics


Marketing

30


Or if you prefer Haiku…

hear mobile marketing sing

a big data song

CMOs will weep with joy

31


Recommended Resources

ANINDYA GHOSEDirector, Center for Business AnalyticsAssociate Professor Stern School of BusinessNew York University

@aghose

MARKETINGSHERPA2012 Mobile Benchmark Reporthttp://bit.ly/UeEIrO

@marketingsherpa

32


JENNIFER VEESENMEYERVice President, Digital Analytics Practice

443.542.4611612.356.4191 (cell)@ @[email protected]

33

mailto:[email protected]

Big Data and Mobile Analytics - MMA SF Jan13

Business

mobile testing

mobile marketing programs

mobile marketing association2

mobile marketing association9

mobile marketing association7

mobile marketing association8

mobile marketing association12

mobile marketing budgetsare