finding meaning in mobile behavioral data - casro webinar

Melanie Courtright and Roddy Knowles

Finding Meaning in Mobile Behavioral Data

2

Can you identify this picture?

3

Notice any differences?

4

Who are we studying?What devices?

5

Male Female Less than $25K

$25K - $49K

$50K - $74K

$75K - $99K

$100K - $149K

$150K - $199K

$200K+

49% 51%

25% 25%

18%

12% 12%

4% 4%

50% 50%

17%20%

23%

14% 15%

6% 5%

Census Mobile

BrandGender Income

Male Female < $25K$25K - $49K

$50K - $74K

$75K - $99K

$100K - $149K

$150K - $199K

$200K+

Android 57% 39%   53% 56% 48% 37% 44% 23% 24%

iPhone 33% 50%   32% 37% 43% 50% 44% 63% 70%

Blackberry 3% 3%   1% 3% 4% 4% 4% 8% 4%

Windows 3% 2%   2% 3% 1% 4% 4% 2% -

Other 5% 7%   12% 3% 4% 5% 4% 4% 3%

US Census vs. Mobile and Brand Demos

6

US Census vs. Mobile and Brand Demos

18 - 24 25 - 34 35 - 44 45 - 54 55+ Hispanic Caucasian Black Asian Other

13%18% 18% 19%

33%

16%

72%

13%

5%9%

13%18% 19% 20%

29%

15%

73%

15%

6% 6%

Census Mobile

BrandAge Ethnicity

18-24 25-34 35-44 45-54 55+ Hispanic Caucasian Black Asian Other

Android 51% 54% 54% 45% 40%   59% 42% 57% 53% 74%

iPhone 43% 42% 36% 39% 45%   28% 47% 30% 35% 19%

Blackberry 2% 1% 2% 5% 5%   5% 3% 8% - 2%

Windows 2% 2% 4% 4% 0%   3% 2% 2% 8% 2%

Other 2% 1% 5% 7% 9%   6% 7% 4% 4% 3%

7

Who We Are StudyingWhat We Can Control!

30K global completes, single sample source, dual mode, with each mode weighted to age, gender, income, and ethnicity.

Sports / Olympics Attitudes Mobile (Mean) Web (Mean)I am only interested in major sporting events 2.95 2.97

I love watching all sports 3.11 3.14

Sports really don’t interest me 2.18 2.17

Watching / engaging in sports is important to me 3.51 3.56

There are more important things in life than sports 3.57 3.56

The Olympics will increase my interest in sports 3.24 3.31

I will watch the Olympics with friends and family 3.66 3.72

There are more important things in the World than the Olympics 3.60 3.59

Olympics Following Methods (%) (%)Television 92% 91%

Newspapers 42% 40%

News website(s) 33% 36%

Radio 31% 32%

Sport website(s) 23% 25%

Mobile phone apps/applications 13% 15%

Smoking (%) (%)Everyday 15.1% 14.2%

Some Days 7.8% 6.5%

Not at all 77.1% 78.1%

8

What Are We Capturing

Location: GPSApps: usage Phone calls: incoming, outgoing, missed, duration Text/SMS Messages: sent, received Email: sent, received Web: URL/websites visitedCamera: usage

Primary Mobile Platforms2

7 Types of Data Collected Passively

9

What We Can Do With…

Location Data (GPS) Considerations

• Not everyone is trackable

• GPS accuracy: Satellite vs. Network

• Type and frequency of communication

• Filter behavioral data by location

• Derive location type (e.g. home, work)

• Target and push surveys

• Advanced analytics

10

What We Can Do With…

App Data Considerations

• Target based on apps used

• Measure “stickiness”

• Slice and dice

• Capture foreground & background usage

• Passive or “active” usage

• Raw data allergies

11

What We Can Do With…

Web/URL Data Considerations

• Track websites/URLs visited

• Aggregate – top sites by category, user

• Compare website usage vs. app usage

• What is captured, what is delivered, and what may be lost

• Time measurement

12

What We Can Do With…

Text/SMS, Email, Phone, Camera, Music

Considerations

• Track ingoing and outgoing communications

• Log Camera usage

• Capture specific details of music played

• Music listening varies by platform

• Boolean capture

• When content is important, ask

13

What This Means for Research

Behavioral data stands alone… or hand in hand with other data points

On its own/ Aggregated

• Key indicators and measures

• Trend spotting

Sliced and diced

• Place and time

• Demographics

• Attitudes and behaviors

Combined with survey data

• Stated vs. actual

• Trust but verify

• Understand the “why”

14

Aggregated Behavioral DataTop Level Report on User Activity

Data from US mobile behavioral panelists rolled into a monthly summary report

15

App UsageDecember

29Average

Number of Apps Used

Per Day

Top 10 App Categories Used

Communications 160%

Social Media 155%

Email 124%

App Stores 100%

Music 80%

Streaming Video 76%

Mapping & Nav. 70%

Shopping 42%

% Reach

4App

Categories w/ 100%+

Reach

16

Website VisitationDecember

Top 10 Websites with the Most Unique Visits

1. google.com

2. facebook.com

3. amazon.com

4. yahoo.com

5. wikipedia.org

6. youtube.com

7. ebay.com

8. espn.go.com

9. accuweather.com

10. walmart.com

Top 10 Most Visited Websites

1. google.com

2. facebook.com

3. amazon.com

4. wikipedia.org

5. youtube.com

6. walmart.com

7. ebay.com

8. answers.yahoo.com

9. twitter.com

10. ups.com

Designates website ranks in top 10 for most visits and unique visits.

17

Behavioral Data AloneHoliday Shopping Study

As part of a broader study on holiday shopping including online and B&M shopping, mobile activity was tracked.

18

Wal

mar

t

Targ

et

Best B

uy

Kohl's

Sear

s

Macy'

s

The

Home

Depot

Costc

o

Toys

"R" U

s

Lowe'

s

Barne

s & N

oble

JC P

enne

y

Sam

's C

lub

Kmar

t

Stap

les

Apple

Sto

re

Nords

trom

Dillar

d's

Gap

North

Face

Other

70%64%

47%

41%35%

33%

24% 23% 22% 22% 21% 21% 20% 18%

10% 10% 9% 8% 7%3%

9%

Planned and actual store visits during a two week period

What respondents said and did ranked differently

2 7 11 3 515 16 101

Top 10 store visitation ranking

GPS Store visitation

0

N=5829 - Q8. Which of the following store locations are you planning to visit in the next few weeks? Please check all that apply.

19

Top 20 Apps Usage

Period: Nov 19-30, 2012. Tracked websites which are in top 20 are highlighted in Blue Color. Segmented based on No of Users & Hrs used is more than 100. Default apps not considered under consideration (like Settings, Contacts etc.,)

The Facebook app was used 11,004 hrs during the 12-day period (11/19-11/30) by 719 users averaging to 1.275 hrs per day (76.5 minutes).

20

Store Visits GPSSample = 800

Mobile Websites Sample = 1150

PC Websites Sample = 5398 To

tal #

of

Vis

its

11/19 11/20 11/21 11/22 11/23 11/24 11/25 11/26 11/27 11/28 11/29 11/30

0

300

242 247 256

136

207

167144 153 144

202

254 245

Tota

l #

of

Vis

its

11/19 11/20 11/21 11/22 11/23 11/24 11/25 11/26 11/27 11/28 11/29 11/30

0

1000

376281 279

456

701745

921

543 547475

204 223

Tota

l #

of

Vis

its

11/19 11/20 11/21 11/22 11/23 11/24 11/25 11/26 11/27 11/28 11/29 11/30

0

7000

3389 33163789

49695404

38854593

5757

3221 3098 2930 2710

BLACK

FRIDAY

CYBER

MONDAYTraffic

21

Behavioral & Survey Data CombinedIn-store Phone Usage

Grocery shoppers’ phone usage was captured, verified, and contextualized.

22

Shopper Profile

Fred Meyer – April 18,

2013

Male, 53, Divorced

Game, Facebook

$14

Special Trip,

Under 15 Minutes

Greg Grocery

No Coupon

Used

23

App UsageStated and Actual

Series1

31%

24%

13% 12%

5%

25%21%

16%14%

11%

Stated Actual

Types of Apps Used

Text/SMS GamesShopping/Retail

Phone Social

24

App UsageWhat & Why

46% Call about a purchase

39%

Message about a purchase

What did you buy?

25

Shopping List

83%

App UsageWhat & Why

45%

41%

Coupon

Compare Prices

26

The Behavioral Data Journey

27

P1: Lost at Sea

28

P2: Learning how the data lives together

29

P3: Happy Days

30

P3b: I Love Data, Too

31

P4: Data Magic

32

P5: Elite Data Team

33

Thank you!

33

Melanie Courtrightmcourtright@researchnow.com

Roddy Knowlesroknowles@researchnow.com