Presentation of the paper " Rise of the Planet of the Apps: A Systematic Study of the Mobile App Ecosystem" at Internet Measurement Conference (IMC) 2013: http://conferences.sigcomm.org/imc/2013/index.html The paper can be found here: http://conferences.sigcomm.org/imc/2013/papers/imc217-petsasA.pdf

1. A Systematic Study of the Mobile App
Ecosystem
Thanasis Petsas, Antonis Papadogiannakis, Evangelos P. Markatos
Michalis Polychronakis Thomas Karagiannis

2. Smartphone Adoption Explodes
• Smartphone adoption:
– 10x faster than 80s PC revolution
– 2x faster than 90s Internet Boom
– 3x faster than social networks
• 1.4 B smartphones in use by 2013!
Source:
2

3. Mobile Apps are Getting Popular
50B+
downloads
1M+
apps
50B+
downloads
900K+
apps
Windows Store
2B+
downloads
100K+
apps 3

4. A Plethora of Marketplaces
• In addition to the official
marketplaces...
• Many alternative markets
4

5. Crawler Hosts
Data Collection
MarketplacesPlanetLab
Proxies
App stats
APKs
App stats
APKs
App stats
APKs
Database
Appstats
APKs
5

6. Datasets
Appstore Crawling
period
Total apps* New apps /
day
Total
downloads*
Daily
downloads
SlideMe (free) 5 months 16,578 28.0 96 M 215.7 K
SlideMe (paid) 5 months 5,606 6.5 914 K 5.2 K
1Mobile 4.5 months 156,221 210.4 453 M 651.5 K
AppChina 2 months 55,357 336.0 2,623 M 24.1 M
Anzhi 2 months 60,196 29.6 2,816 M 23.7 M
* Last Day
~ 300K apps
Paid apps:
• less downloads
• fewer uploads
6

7. App Popularity
Is There a Pareto Effect?
Downloads(%)CDF
Normalized App Ranking (%)
7

8. App Popularity
Is There a Pareto Effect?
Downloads(%)CDF
Normalized App Ranking (%)
10% of the apps account for
90% of the downloads
7

9. App Popularity
Is There a Power-law Behavior?
8

10. App Popularity
Is There a Power-law Behavior?
Let’s focus on
one appstore
8

11. App Popularity
Deviations from ZIPF
9

12. App Popularity
Deviations from ZIPF
WWW
INFOCOM‘99
9

13. App Popularity
Deviations from ZIPF
WWW
INFOCOM‘99
9

14. App Popularity
Deviations from ZIPF
WWW
INFOCOM‘99
P2P
SOSP’03
9

15. App Popularity
Deviations from ZIPF
WWW
INFOCOM‘99
P2P
SOSP’03
UGC
IMC’07
9

16. Truncation for small x values:
Fetch-at-most-once
• Also observed in P2P workloads
• Users appear to download an
application at most once
P2P
SOSP’03
simulations
10

17. Truncation for large x values:
clustering effect
• Other studies attribute this truncation to information filtering
• Our suggestion: the clustering effect
UGC
IMC’07
11

18. App Clustering
• Apps are grouped into clusters
• App clusters can be formed by
– App categories
– Recommendation systems
– User communities
– Other grouping forces
12

19. Clustering Hypothesis
• Users tend to download apps from the same
clusters
I like
Games!
I like Social
apps!
13

20. Validating Clustering Effect in User
Downloads
Dataset: 361,282 user comment streams,
60,196 apps in 34 categories
14

21. Validating Clustering Effect in User
Downloads
Dataset: 361,282 user comment streams,
60,196 apps in 34 categories
53% of users commented
on apps from a single category
14

22. Validating Clustering Effect in User
Downloads
Dataset: 361,282 user comment streams,
60,196 apps in 34 categories
94% of users commented
on apps from up to 5 categories
14

23. User Temporal Affinity
a1 a2 a3 a4 a5
User downloads
sequence
a1, a2, a3, a4, a5
x
Aff1 = 0
✔
Aff2 = 1
✔
Aff3 = 1
x
Aff4 = 0
𝐴𝑓𝑓𝑖𝑛𝑖𝑡𝑦 =
Affi
𝑛−1
𝑖=1
𝑛 − 1
=
0 + 1 + 1 + 0
4
= 0.5
15

24. Users Exhibit a Strong Temporal
Affinity to Categories
0.55
0.14
16

25. Users Exhibit a Strong Temporal
Affinity to Categories
0.55
0.14
3.9 x
16

26. Modeling Appstore Workloads
. . .
Top
bottom
Apppopulatiry
ReaderGames Social Productivity
APP-CLUSTERING model
17

27. Modeling Appstore Workloads
. . .
Top
bottom
Apppopulatiry
ReaderGames Social Productivity
APP-CLUSTERING model
1. Download the 1st app – overall app ranking
1
17

28. Modeling Appstore Workloads
. . .
Top
bottom
Apppopulatiry
ReaderGames Social Productivity
APP-CLUSTERING model
1. Download the 1st app – overall app ranking
17

29. Modeling Appstore Workloads
. . .
Top
bottom
Apppopulatiry
ReaderGames Social Productivity
APP-CLUSTERING model
1. Download the 1st app – overall app ranking
2. Download another app
2
17

30. Modeling Appstore Workloads
. . .
Top
bottom
Apppopulatiry
ReaderGames Social Productivity
APP-CLUSTERING model
1. Download the 1st app – overall app ranking
2. Download another app
2.1 with prob. p from a previous app cluster c – cluster app ranking
2.1
p
17

31. Modeling Appstore Workloads
. . .
Top
bottom
Apppopulatiry
ReaderGames Social Productivity
APP-CLUSTERING model
1. Download the 1st app – overall app ranking
2. Download another app
2.1 with prob. p from a previous app cluster c – cluster app ranking
17

32. Modeling Appstore Workloads
. . .
Top
bottom
Apppopulatiry
ReaderGames Social Productivity
APP-CLUSTERING model
1. Download the 1st app – overall app ranking
2. Download another app
2.1 with prob. p from a previous app cluster c – cluster app ranking
2.2 with prob. 1-p – overall app ranking
2.2
1-p
17

33. Modeling Appstore Workloads
. . .
Top
bottom
Apppopulatiry
ReaderGames Social Productivity
APP-CLUSTERING model
1. Download the 1st app – overall app ranking
2. Download another app
2.1 with prob. p from a previous app cluster c – cluster app ranking
2.2 with prob. 1-p – overall app ranking
17

34. Modeling Appstore Workloads
. . .
Top
bottom
Apppopulatiry
ReaderGames Social Productivity
APP-CLUSTERING model
1. Download the 1st app – overall app ranking
2. Download another app
2.1 with prob. p from a previous app cluster c – cluster app ranking
2.2 with prob. 1-p – overall app ranking
3. If user’s downloads < d go to 2.
If downloaded apps < user downloads
 go to 2.
3
17

35. Model Parameters
Symbol Parameter Description
A Number of apps
D Total downloads
d Downloads per user (average)
C Number of clusters
U Number of users
zr Zipf exponent for overall app ranking
ZG Overall Zipf distribution of all apps
P Percentage of downloads based on clustering effect
zc Zipf exponent for cluster’s app ranking
Zc Zipf distribution of apps in cluster c
D(I,j) Predicted downloads for app with total rank i and
rank j in its cluster 18

36. Model Parameters
Symbol Parameter Description
A Number of apps
D Total downloads
d Downloads per user (average)
C Number of clusters
U Number of users
zr Zipf exponent for overall app ranking
ZG Overall Zipf distribution of all apps
P Percentage of downloads based on clustering effect
zc Zipf exponent for cluster’s app ranking
Zc Zipf distribution of apps in cluster c
D(I,j) Predicted downloads for app with total rank i and
rank j in its cluster 18

37. Model Parameters
Symbol Parameter Description
A Number of apps
D Total downloads
d Downloads per user (average)
C Number of clusters
U Number of users
zr Zipf exponent for overall app ranking
ZG Overall Zipf distribution of all apps
P Percentage of downloads based on clustering effect
zc Zipf exponent for cluster’s app ranking
Zc Zipf distribution of apps in cluster c
D(I,j) Predicted downloads for app with total rank i and
rank j in its cluster
 Number of downloads of
the most popular app
18

38. Results
AppChina
19

39. Results
AppChina
19

40. Results
AppChina
19

41. Results
AppChina
19

42. Results
AppChina
Model Distance from
measured data
ZIPF 0.77
ZIPF-at-most-once 0.71
APP-CLUSTERING 0.15
19

43. App Pricing
• Main Questions:
– Which are the differences between paid & free apps?
– What is the developers’ income range?
– Which are the common developer strategies
• How do they affect revenue?
20

44. The influence of cost
Free Paid
21

45. The influence of cost
Clear
power-law
Free Paid
21

46. The influence of cost
Clear
power-law
Free Paid
Users are more selective when downloading paid apps
21

47. Developers’ Income
22
(USD)

48. Developers’ Income
Median: < 10 $
22
(USD)

49. Developers’ Income
80% < 100 $
22
(USD)

50. Developers’ Income
95% < 1500 $
22
(USD)

51. Developers’ Income
22
(USD)

52. Developers’ Income
22
(USD)

53. Developers’ Income
Quality is more important than quantity
22
(USD)

54. Developers Create a Few Apps
23

55. Developers Create a Few Apps
A large portion of developers
create only 1 app
23

56. Developers Create a Few Apps
95% of developers
create < 10 apps
23

57. Developers Create a Few Apps
10% of developers offer
free & paid apps
23

58. Can Free Apps Generate Higher Income
Than Paid Apps?
Necessaryadincome(USD)
Day
24

59. Can Free Apps Generate Higher Income
Than Paid Apps?
Necessaryadincome(USD)
Day
Average: 0.21 $
24

60. Can Free Apps Generate Higher Income
Than Paid Apps?
Necessaryadincome(USD)
Day
Average: 0.21 $
An average free app needs about
0.21 $/download to match the income of a paid app
24

61. Conclusions
• App popularity: Zipf with truncated ends
– Fetch-at-most-once
– Clustering effect
• Practical implications
– New replacement policies for app caching
– Effective prefetching
– Better recommendation systems
– Increase income
25

62. Thank you!
26