Clustering by mobile usage and behaviour – the many faces of smartphone users in South Africa - GfK

November 5 & 6, 2014
DoubleTree by Hilton Hotel Cape Town
Upper Eastside, South Africa
#MRMW
Organized by

#MRMW
Title Sponsor Silver Sponsors Bag Sponsor
Association & Media Partners

CLUSTERING BY MOBILE USAGE AND BEHAVIOUR – THE MANY
FACES OF SMARTPHONE USERS IN SOUTH AFRICA
Ryan Smit
GfK Digital Market Intelligence

2© GfK 2014 | GfK Digital Market Intelligence
GfK
With headquarters in Nuremburg, Germany, GfK is
present in over 100 countries around the world. We
have been trusted for over 80 years.
Our African footprint extends into 36
countries and some of the biggest economies
in Africa.
We have experts in every major industry to answer
our clients’ key market questions and the
techniques to understand what is happening
where and why

TRACKING CONSUMER MOBILE BEHAVIOUR,
ANONYMOUSLY AND PASSIVELY

GfK has created a mobile application called FUNDI in sub-
Sahara Africa to build a community of consumers who can
benefit from free content related to education, business &
entrepreneurship – something that we know is close to
their hearts.
This means access to the communities’ smartphone usage
data (via GfK’s proprietary LEOtrace technology which users
download as part of the app) linked to demographic
information: Internet as well as app usage, location data and
much more.
Understanding sub-Sahara African customers via their
mobile phone!
This, of course gives you with the ability to see and analyze
behavioral patterns.
The FUNDI Application

Business & entrepreneurial content is popular among South African of all
backgrounds - according to AMPS, only 15% of the readers of Entrepreneur
Magazine are self employed.
Even when comparing the consumptions patterns between consumers who
are self-employed (entrepreneurs) and those who are not, there are no
significant differences in mobile usage patterns.
It should also be noted that FUNDI is not targeted specifically at
entrepreneurs, but any consumers who are interested in business, education
and entrepreneurship.
Source AMPS 2013 (n=22k)
Self-employed
Use Cellphone for: Yes No
Banking 6% 4%
Chat 19% 18%
Dating 2% 3%
Directory 3% 3%
Apps 6% 7%
Podcast 2% 2%
Email 8% 8%
Gambling 1% 1%
Games 8% 10%
Instant Msg 20% 18%
Radio 4% 5%
Music 10% 12%
Search 12% 14%
Social Networking 14% 15%
Why choose business and entrepreneurship as the theme of the
application?
No part of this document may be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior written permission of GfK South Africa

What kind of data points do we measure for smartphone app
users?
How?What?Who?
Structural Information
• Socio-demographics from
community registration
form (to include age and
SEC)
Apps & Internet
• Application Usage
• Browser Search Queries
• Browser URLs
Data Traffic
• 3g vs Wi-Fi
• Upload vs Download
Communication (Q4 2014)
• Calls
• SMS
Network (Q4 2014)
• Network Provider
• Connection Type
• Signal Strength
• Tower ID & Location
Device
• Model
• Manufacturer
• Operating System Version
Potential Linked Data
Points
• Display Size
• Processor
• CPU Cores
• RAM
• 2G/3G/4G
• Storage (GBs)
• Memory (MBs)
Where and
when?
Time
• Local Time
• UTC time
Location (Q4 2014)
• Longitude/Latitude
• Accurancy and source

Period:
Data Collected & Analysed
May June July
Sample Size:
Data Collected:
15 million records, 200 million data points!
1138 Users

FUNDI Data Weighting Process Overview
NETWORK
PROVIDER
QUARTERLY
REPORTS*
AMPS 2013 Data
– Filtered by those who
have a cell phone, and
report using data
dependant activities
(Apps, Browser etc.)
Extracted
#of
Smartphones
in South Africa
Segmented
by
UNIVERSE
(# of
Smartphones, by
Gender, Age, Race
and Province,
Device Type)
Compared
to…
FUNDI PANEL
Demographics by
Gender, Age, Race
and Province,
Device Type.
Panellist
Weightings
TIME
each panellist
has been in the
in panel during
reporting
period

On average, consumers
spend 117 minutes of
every day on their
smartphones.
Time spent on a
smartphone increases
throughout the day from
around 3 minutes per
hour in the mornings, and
peaks during the evenings
to almost 8 minutes per
hour.
Total time spent on smartphone – by time of the day
Source: GfK SA Smartphone Usage Data – May-July 2014 – Sample Size: 1138 - Usage is defined as the total amount of time their device is active (not in stand-by mode) excluding
home screen usage.
0
1
2
3
4
5
6
7
8
9
Minutesperhour

The average female
consumer spends 122 min
of her day on her
smartphone, while males
spend an average of 113
minutes on their devices.
The usage patterns
throughout the day are
fairly similar, however
females spend an even
greater amount of time on
their devices between 20:00
and 22:00 as compared to
males.
Total time spent on smartphone – by time of the day &
gender
Source: GfK SA Smartphone Usage Data – May-July 2014 – Sample Size: 545 - Usage is defined as the total amount of time their device is active (not in stand-by mode).
0
1
2
3
4
5
6
7
8
9
10
Minutesperhour
Female Male

In terms of share of usage,
the most time, by far, is
spent on communication
activities (44%) followed by
Social Activities (15%),
Gaming (15%), Browsing
(8%) and Media & Video
(5%)
These combined account
for 87% of total time spent
on a smartphone, with the
other 13% split among the
remaining categories.
Share of time spent on smartphone by application
category
Source: GfK SA Smartphone Usage Data – May-July 2014 – Sample Size: 1138 – Share of time is measured as the total time spent using those applications during the reporting period.
Categories are based on Android Play Store categorization. Categories excluded: Tools, Personalization, Libraries & Demo, Live Wallpapers
44%
15%
15%
8%
5%
2%
2%
1%
1%
1%
1%
1%
1%1% 1% 1%
Communication
Social
Games
Browsing
Media & Video
Music & Audio
Photography
News & Magazines
Books & Reference
Travel & Local
Productivity
Sports
Entertainment
Finance
Lifestyle
Other

13© GfK 2012 | GfK Digital Market Intelligence | 14. February 2012
WHAT CAN MOBILE BASED CLUSTERS
TELL US ABOUT CONSUMERS?

Cluster Analysis Methodology
Cluster analysis seeks to
identify natural clusters in
a population by
maximising the similarities
within clusters, and
maximising the difference
between clusters.
In order to identify
behavioural differences,
our data was clustered
using a combination of
the 5 most used
application categories…

Input variables, and examples per application category
Communications
•WhatsApp
•BBM
•Phone
•Messaging
•Email
•etc.
Social
•Facebook
•Instagram
•Twitter
•Google+
•Pinterest
•etc.
Games
•Candy Crush Saga
•Farm Heroes Saga
•Clash of Clans
•Hay Day
•My Talking Tom
•etc.
Browsing
•Internet Browser
•Chrome Browser
•Opera Mini
•Firefox Browser
•UC Browser
•etc
Media & Video
•Gallery
•YouTube
•VLC for Android Beta
•Movies
•MX Player
•etc.

So what clusters emerged?
7 segments emerged from
the data, each with specific
characteristics which
differentiate them from
other segments of
smartphone users.
So what are the segments,
and what can they tell
us about smartphone
users in South
Africa?
Surfers
(7%)
Heavy
Gamers
(15%)
Heavyweights
(19%)
Video
Gamers
(7%)
Lightweights
(29%)
Chatters
(14%)
Socialites
(9%)

R15k-R40k R40k+
38% 14%

R15k-R40k R40k+
40% 8%
24.3%

R15k-R40k R40k+
37% 5%

R15k-R40k R40k+
11% 5%

R15k-R40k R40k+
37% 9%

R15k-R40k R40k+
30% 4%

R15k-R40k R40k+
17% 2%

LINKING CLUSTERS TO APPLICATION USE:
COMPARING USAGE PATTERNS OF THE TOP
SOCIAL MEDIA PLATFORMS

Usage of Social Networking by Cluster
0.00
5.00
10.00
15.00
20.00
25.00
30.00
Minutesperday
Usage of Top 5 mobile social networks across clusters
Facebook Instagram Twitter GooglePlus LinkedIn
As would be explained
previously, Socialites are
most active on Social
Networking, and this
applies across all social
networks.
Heavyweights are also
extreme users of social
networking, including both
Instagram and Twitter, the
2nd and 3rd most used
social networks on
smartphones.
Source: GfK SA Smartphone Usage Data – May-July 2014 – Sample Size: 1138

Socialites – Usage of social networking over a day
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
Minutesperhour
Usage of top 5 social networks over the course of a day
Facebook GooglePlus Instagram Twitter LinkedIn
Socialites spend most of
their time on Facebook
throughout the day, but
over time they increase
their usage of both
Instagram and Twitter to
peak in the evening
between 18:00 and 22:00.
There is another small peak
of Facebook usage in the
morning, as users check
their News Stream just
after waking up.
Source: GfK SA Smartphone Usage Data – May-July 2014 – Sample Size: 1138

FUTURE EVOLUTION OF CLUSTERS

Future evolution of the smartphone clusters……
Activity Based
• Communication
• Social Networking
• Wikipedia
• Games & Media
• Tinder
Needs Based
• Interaction
• Community
• Information
• Leisure
• Love and Intimacy

THANK YOU

Clustering by mobile usage and behaviour – the many faces of smartphone users in South Africa - GfK

More Related Content

What's hot (19)

Viewers also liked (15)

Similar to Clustering by mobile usage and behaviour – the many faces of smartphone users in South Africa - GfK (20)

More from Merlien Institute (20)

Recently uploaded (20)

Clustering by mobile usage and behaviour – the many faces of smartphone users in South Africa - GfK