Abstract
Understanding exactly how consumers are using their smartphones presents a
greater challenge than previous generations of handset.
|
|
We used on-device trackers to gather device usage data from a panel of
consumer smartphone users in France, Germany, Spain, the UK and the USA during
August and September 2011. In this Report, we consider how consumers use their
smartphones and the impact that this has on cellular data traffic, as well as
Wi-Fi connectivity.
|
This Report provides answers to the following questions.
- How much data are they generating and over which network?
- Which apps are generating data and how does the quantity of that data vary
between Wi-Fi and cellular networks?
- Who are the consumers who do not use cellular data on their handset and
are OTT apps cannibalising core voice and messaging spend for these customers?
- Are non-data-using consumers worth investing in? How can they be
encouraged to use data?
- Who are the heaviest smartphone users?
- What role do handset types and Wi-Fi have to play in the volume of traffic?
- How do consumers use Wi-Fi hotspots with their smartphone?
- What type of hotspots do they use, and how many hotspots?
- How does usage behaviour change when consuemrs are abroad and which
customers are operators capturing/losing?
Figure 36: The distribution of per-minute data traffic generated by
web browsers on Android and iOS-based smartphones
[Source: Analysys Mason and Arbitron Mobile, 2012]
1 n = 1007; some respondents may have had cellular or
Wi-Fi enabled but did not use the connectivity within the observation period.
About the author
Martin Scott (Principal Analyst) co-ordinates Analysys Mason's primary
research report series, including reports related to consumer smartphone usage
and the Connected Consumer series. Martin also leads Analysys Mason's Fixed
Broadband research programme and contributes regularly to the Mobile Broadband
and Devices programme. His primary areas of specialisation include customer
satisfaction and consumer-facing marketing strategy, broadband retail pricing
and bundling. Martin also specialises in statistics, surveys and the analysis
of primary research. He has produced research for Analysys Mason on different
aspects consumer demand for present and next-generation services, the business
case for value-added services, such as videotelephony and three-screen
advertising and broadband (next-generation) access. Martin has a Master's
degree in Mathematics from Oxford University.
Table of Contents
- 6. Executive summary
- 7. Understanding consumer smartphone data traffic and network usage
through on-device measurement
- 8. Mobile video is not the killer app (yet): the browser generates the
greatest quantity of smartphone Wi-Fi data traffic and email dominates cellular
- 9. 18% of smartphone users in our panel did not use cellular data: OTT
communication apps may begin to erode core revenue in this segment
- 10. A minority of heavy users distort total traffic levels: these are
almost all iPhone users
- 11. Recommendations
- 12. Implications and recommendations
- 13. Implications and recommendations
- 14. Real-world usage: we measured consumer smartphone usage via an
on-device monitoring application, in partnership with Arbitron Mobile
- 15. The two leading operating systems - Android and iOS - are
over-represented in the smartphone user panel
- 16. Key questions and issues discussed in this report
- 17. Wi-Fi versus cellular: gaining a full picture of smartphone data
usage
- 18. On-device measurement offers a clear view of the split between
cellular and Wi-Fi traffic among our smartphone user panellists
- 19. Cellular data users in the USA are slightly ‘hungrier' than in
France, Germany and the UK
- 20. Younger consumers generate more smartphone data traffic on cellular
and Wi-Fi networks than older consumers
- 21. Connectivity is linked to usage: Wi-Fi-only panellists were light
users, while panellists who used both Wi-Fi and cellular were heavier users
- 22. Half of our smartphone panellists generated less than 221MB of
smartphone data (across both cellular and Wi-Fi) per month
- 23. Wi-Fi had a greater number of very light and very heavy users whereas
cellular data usage patterns are distributed more regularly
- 24. More than 40% of Android-using panellists connected to two or more
Wi-Fi hotpots, and almost a third connected to public hotspots
- 25. More panellists used roaming cellular data than Wi-Fi hotspots when
travelling abroad
- 26. Apps and traffic: identifying the apps that drive data traffic
- 27. Our analysis of traffic generated by specific apps indicates that rich
media - in particular video - is not driving the majority of smartphone traffic
- 28. The combination of browser and email apps on a smartphone generated
almost four times the quantity of data traffic as social networking
- 29. The browser generates the greatest amount of smartphone Wi-Fi traffic,
whereas email dominates cellular data use
- 30. Cellular data traffic peaks at 09.00 and 19.00, whereas Wi-Fi traffic
builds to 22.00
- 31. Non-cellular data users: understanding Wi-Fi-only smartphone
users
- 32. 18% of smartphone users in our panel did not use cellular data
- 33. Consumers who own a smartphone but do not use cellular data are not
demographically easy to identify
- 34. Android users are less likely to use cellular data than iPhone users,
and US panellists are less likely to use cellular data than European panellists
- 35. Smartphone users who do not use cellular data tend to be either light
or heavy voice users; heavy voice users should be a key target for selling data
- 36. Wi-Fi-only smartphone users often use OTT communications apps: this
may erode spending on the core services of SMS and voice
- 37.‘Power users': traffic and usage among the heaviest data
users
- 38. 1% of panellists generated almost 20% of traffic in the study: such
consumers should be moderated or monetised
- 39. iPhone users are heavier users of data than other consumers, and
increasing iPhone penetration could stimulate data revenue growth
- 40. Heavier cellular data users are not necessarily also heavier Wi-Fi
users - ‘power users' are not that common
- 41. Operators' different approaches to data allowances appear to have a
significant affect on the number of heavy cellular data users in a country
- 42. Web browsing on an iPhone generates twice the per-minute traffic of
browsing on an Android smartphone
- 43. Methodology and definitions
- 44. Methodology and definitions [1]
- 45. Methodology and definitions [2]
- 46. Methodology and definitions [3]
- 47. About Arbitron Mobile
- 48. About the author and Analysys Mason
- 49. About the author
- 50. About Analysys Mason
- 51. Research from Analysys Mason
- 52. Consulting from Analysys Mason
List of figures
- Figure 1: Distribution of total smartphone traffic across all panellists
- Figure 2: Relative traffic volumes generated by different apps, by network
- Figure 3: Distribution of smartphone panellists, by type of data
connectivity
- Figure 4: The average monthly data usage of the ten heaviest individual
data users in the panel of smartphone users
- Figure 5: Illustration of Analysys Mason - Arbitron smartphone data
analysis process
- Figure 6: Smartphone panellists included in the data traffic analysis, by
country
- Figure 7: Smartphone panellists included in the data traffic analysis, by
age
- Figure 8: Smartphone panellists included in the data traffic analysis, by
OS
- Figure 9: Structure of this report and key issues addressed
- Figure 10: Distribution of total smartphone traffic in the panel
- Figure 11: Average monthly cellular data consumption, by country
- Figure 12: Average monthly cellular and Wi-Fi smartphone data consumption,
by age range
- Figure 13: Panellists' average smartphone data usage
- Figure 14: Distribution of total average monthly smartphone data traffic
for panellists, by percentile
- Figure 15: Distribution of total monthly smartphone cellular data traffic,
by respondent's data usage percentile
- Figure 16: Distribution of total monthly smartphone Wi-Fi traffic, by
percentile
- Figure 17: Distribution of Wi-Fi hotspot usage for Android users during
the two-month observation period
- Figure 18: Percentage of panellists who used Wi-Fi that connected to
different types of Wi-Fi hotspot
- Figure 19: Types of data connectivity used abroad by panellists who spent
any time away from their home country during the observation period
- Figure 20: Data traffic generated by smartphone panellists, by foreground
or background apps
- Figure 21: Relative volumes of smartphone data traffic generated by
different apps
- Figure 22: Relative volumes of smartphone data traffic generated by
different apps, by network
- Figure 23: Cellular and Wi-Fi data volumes as a percentage of total data
traffic generated, by time of day
- Figure 24: Distribution of smartphone panellists, by type of data
connectivity
- Figure 25: Percentage of panellists who used different data connectivity
options, by age, relative to the total panel
- Figure 26: Percentage of panellists who used different data connectivity
options, by gender, relative to the total panel
- Figure 27: Percentage of panellists who used different data connectivity
options, by country, relative to the total panel
- Figure 28: Percentage of panellists who used different data connectivity
options, by OS, relative to the total panel
- Figure 29: Distribution of average monthly outgoing voice minutes, by data
usage profile
- Figure 30: Distribution of average monthly SMS sent, by data usage profile
- Figure 31: Percentage of panellists who use OTT communications apps, by
type of data connectivity
- Figure 32: The average monthly data usage of the ten heaviest data users
in the smartphone panel
- Figure 33: Distribution of monthly data traffic percentiles by smartphone
operating system
- Figure 34: The relationship between the relative heaviness of Wi-Fi and
cellular data usage by panellist percentile
- Figure 35: Demographic breakdown of the top-10% heaviest cellular data
users relative to the total panel
- Figure 36: The distribution of per-minute data traffic generated by web
browsers on Android and iOS-based smartphones
- Figure 37: Utility app categorisation examples
- Figure 38: Utility app categorisation examples
