iPhone AppStore Secrets - Pinch Media

1. AppStore
Secrets’
 (What
We’ve
Learned
From
30,000,000
Downloads)

 Greg
Yardley
 Co‐Founder
&
CEO
 greg@pinchmedia.com
 646‐330‐8540

2. 30,000,000
Downloads?!
 (Actually,
it’s
a
fair
bit
more
than
that
by
now.)
 •  Since
AppStore
launch,
Pinch
Media
has
provided
 developers
with
an
analyUcs
library
to
monitor
app
usage
 –
unique
users,
sessions,
usage
Ume,
etc.

 •  Since
AppStore
launch
we’ve
also
been
collecUng
every
 bit
of
detail
possible
from
the
AppStore
–
rankings,
price
 changes,
you
name
it
–
and
tying
it
back
to
our
analyUcs.
 •  Our
stuff’s
in
a
few
hundred
applicaUons
right
now
–
it’s
 been
in
the
#1
free
and
paid
applicaUon
several
Umes
 each,
and
has
been
in
at
least
ten
of
the
top
100
free
 applicaUons
for
a
while
now.
 •  With
all
of
this
data,
you
learn
a
few
things.



3. ApplicaUon
Rankings
 (How
does
the
AppStore
work,
anyway?)
 For
every
ranked
list
on
the
AppStore,
here’s
a
good
 rule
of
thumb:
 24‐hour

 rolling
window

 of
units
downloaded
 (So
bunch
up
your
publicity.)

4. What
do
you
get
by
appearing
on
a
list?
 •  Appearing
on
a
top
100
list
increases
daily
 new
users
by
an
average
of
2.3x.
 •  Greater
gains
result
from
appearing
in
the
top
 25
and
top
10
lists
–
more
variable,
but
oaen
 an
order
of
magnitude.
 •  However,
it’s
not
permanent.

Apple’s
 AppStore
is
structured
for
maximum
turnover.

5. Case
Study
A:
Well‐Timed
Price
Cut

6. Case
Study
B:
Not‐So‐Well
Timed

7. Case
Study
C:
CounterproducUve?

8. In
general…
 •  Don’t
mess
with
a
posiUve
download
trend.
 •  Decreasing
price
is
oaen
worthwhile.
 •  Aaer
you’ve
been
broadly
exposed,
 experiments
have
less
effect.

 The
average
price
cut
increased
demand
by
130%.
 The
average
price
increase
drops
demand
to
25%.

9. What
do
I
need
to
get
on
a
list?
 For
free
applicaUons:
 Top 25 Top 100 six months ago 10,000 1,000 three months ago 11,000 1,500 today 20,000 5,000 (Apple
had
a
big
Christmas!)

10. Case
Study
D:
Happy
Holidays

11. Do
I
have
a
community?
 (aka
‘How
much
is
my
app
used?’)
 •  So
you’ve
got
a
million
downloads
–
congrats!

 But
what
percentage
use
your
applicaUon
the
 next
day?

The
day
aaer?
 •  The
biggest
applicaUons
in
our
system
have
 +3MM
downloads
–
but
what
kind
of
acUve
 user
base
does
a
download
translate
into?

12. Free
ApplicaHons
‐
Usage
Over
Time
 25.00%
 Users
Returning
(%
of
Day
0)
 20.00%
 15.00%
 10.00%
 5.00%
 0.00%
 1
 11
 21
 31
 41
 51
 61
 71
 81
 91
 Days
Since
First
Used

13. Paid
ApplicaHons
‐
Usage
Over
Time
 35.0%
 Users
Returning
(%
of
Day
0)
 30.0%
 25.0%
 20.0%
 15.0%
 10.0%
 5.0%
 0.0%
 1
 11
 21
 31
 41
 51
 61
 71
 81
 91
 Age
Since
First
Used

14. ApplicaHons
By
Category
‐
Usage
Over
Time
 30.0%
 Entertainment
 Users
Returning
(%
of
Day
0)
 Games
 25.0%
 Sports
 Lifestyle
 20.0%
 UUliUes
 15.0%
 10.0%
 5.0%
 0.0%
 1
 4
 7
 19
 25
 28
 31
 37
 49
 55
 58
 61
 67
 79
 85
 88
 91
 97
 100
 10
 13
 16
 22
 34
 40
 43
 46
 52
 64
 70
 73
 76
 82
 94
 Days
Since
First
Use

15. In
other
words…
 •  Users
stop
using
the
average
applicaUons
 prely
quickly.

Long‐term
audiences
are
 generally
1%
of
total
downloads.
 •  Paid
applicaUons
generally
retain
their
users
 longer
than
free
applicaUons,
although
the
 drop‐off
is
sUll
prely
steep.
 •  Sports
seems
beler
at
retaining
users
over
 the
short
term;
entertainment
at
retaining
 users
over
the
long
term.



16. How
long
are
they
using
it?

 •  For
certain
applicaUons,
the
length
of
Ume
 users
use
the
applicaUon
is
important.

 •  Branded
applicaUons
care
deeply
about
 engagement.
 •  ApplicaUons
showing
ads
periodically
also
care
 about
session
length,
for
obvious
reasons.
 •  In
general,
every
second
the
app’s
open
is
a
 second
it
can
be
seen
by
or
recommended
to
 others.



17. So
should
I
give
it
away
or
not?
 •  Anyone
browsing
the
top
free
applicaUons
 knows
that
adverUsing
is
an
opUon.
 •  The
biggest
player
is
AdMob,
but
Pinch
Media
 has
some
partnerships
with
ad
networks
that
 supply
some
of
these
ads.

 •  However…
I
used
to
be
much
more
 enthusiasUc
about
adverUsing
than
I
am
today.

 Here’s
why:

18. Total
ApplicaHon
Runs
Since
First
Use
 12
 10
 Total
ApplicaHon
Runs
 8
 6
 4
 2
 0
 1
 11
 21
 31
 41
 51
 61
 71
 81
 Days
Since
First
Use

19. Average
‘free
vs.
paid’
raUos:
 •  for
total
unique
users:
 7.5
to
1
 •  for
total
number
of
Umes
used:
 6.6
to
1
 •  for
total
Ume
spent
using
the
applicaUon:
 3.9
to
1



20. ExtrapolaUng…
 •  Assume
free
applicaUons
are
run,
at
most,
a
 dozen
Umes
per
user.
 •  We
see
free
applicaUons
run,
on
average,
6.6
 Umes
as
oaen
as
paid
applicaUons.
 •  A
paid
applicaUon
returns
at
least
$0.70
/
user.
 •  Doing
the
math
–
12
x
6.6
=
80
sessions.
 •  Can
the
average
applica/on
make
more
than
 $0.70
off
adver/sing
in
80
sessions?




21. Answer:
Hell
no.
 Earning
$0.70
in
80
sessions
requires
revenue
of
 $8.75
per
thousand
runs.
 If
you
can
show
one
ad
per
session,
that’s
an
 $8.75
CPM.
 Right
now,
with
the
ad
market
how
it
is,
 adverUsing
rates
of
$0.50‐$2.00
CPM
are
 much
more
typical.
 The
typical
applicaUon
would
have
to
bombard
 its
users
with
ads
to
beat
the
money
it’d
make
 from
paid
sales.





22. But
adverUsing
isn’t
always
a
bad
idea.

 •  Some
applicaUons
benefit
from
network
 effects,
and
get
far
more
than
6.6x
the
users
 they’d
get
if
they
charged.
 •  Some
applicaUons
are
excepUonally
‘sUcky’
–
 users
use
the
app
far
more
than
average.
 •  Some
applicaUons
–
generally,
ones
catering
 to
people
with
money
–
can
command
beler
 adverUsing
rates
than
usual.


23. CumulaHve
ApplicaHon
Runs
Since
First
Use,
By
Decile
 45
 40
 35
 30
 ApplicaHon
Runs
 25
 20
 15
 10
 5
 0
 1
 6
 11
 16
 21
 26
 31
 36
 41
 46
 51
 56
 61
 66
 71
 76
 81
 86
 91
 96
 Days
Since
First
Used

24. CumulaHve
ApplicaHon
Runs
Since
First
Use,
By
Decile
 CPM
 45
 <
$2.00
 40
 35
 30
 ApplicaHon
Runs
 25
 20
 15
 ~
$7.00
 10
 ~
$15.00
 5
 ~
$35.00
 0
 1
 6
 11
 16
 21
 26
 31
 36
 41
 46
 51
 56
 61
 66
 71
 76
 81
 86
 91
 96
 Days
Since
First
Used

25. To
sum
up…
 •  Only
a
few
(<5%)
high‐performing
applicaUons
 are
suitable
for
adverUsing
right
now,
and
you
 don’t
know
if
you’ve
got
one
unUl
aaer
launch.

 •  In
other
words
‐
unless
there’s
something
 inherent
about
the
app
that
screams
free,
sell
it.
 •  Install
analyUcs
in
your
applicaUon
and
watch
 your
sessions
per
user
over
Ume.

Within
a
few
 weeks,
you’ll
know
if
you’ve
got
a
sUcky
 applicaUon.
 •  Only
release
an
ad‐supported
version
when
you
 have
data
strongly
indicaUng
success.

26. Again,
summing
up
‐
 •  Usage
Ume
declines
by
almost
a
third
in
the
 first
month
aaer
use,
stabilizing
at
just
under
 five
minutes.

 •  Paid
applicaUons
see
slightly
more
use
soon
 aaer
installaUon,
and
are
used
for
slightly
 longer
periods.
 •  The
biggest
usage
differenUator
is
category
–
 games
are
used
for
longer
periods
than
any
 other
type
of
applicaUon.

27. This
was
actually
a
sneak
preview
 •  AppStore‐wide
reports
are
being
generated
 daily
and
will
be
incorporated
into
Pinch
 Media’s
reporUng
site
in
the
near
future.
 •  Any
applicaUon
using
our
analyUcs
library
and
 acUvely
sending
in
data
gets
access
to
all
 ecosystem‐wide
reporUng
for
free.

 •  Pinch
Media
wants
to
know
what
else
you
 want
baked
into
this
reporUng.


Firefox 4: fast, powerful and empowering

 Firefox 4: fast, powerful and empowering

1. Firefox 4 1
2. Wait - what happened to Firefox 3.7? • Firefox 3.7 primarily motivated by out of process plugins • UX team suggested longer “bake time” for theme update • Jetpack reboot removed dependency pressures • New engineering capabilities allowed for back-porting • We turned Firefox 3.7 into Firefox 3.6.4 2
3. Wait - what happened to Firefox 3.7? 3.6 3.6.1 3.6.3 3.6.4 Firefox 3.6 OOPP The primary goal for Firefox 3.7 was out of process plugins, which will be shipped to users in Firefox 3.6.4 Firefox.next A1 A2 A3 A4 Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec 3
4. Firefox 4 4
5. Firefox 4 - Motivation 1. Who uses Firefox? 2. What are those people trying to do? 3. How can Firefox be the best tool for the job? 4. What does Firefox need to succeed? 5. How can Firefox help lead the Open Web forward? 5
6. Firefox 4 - Motivation • Who uses Firefox? • End users (early adopters, mainstream) • Web developers (early adopters, mainstream) • Firefox developers (wait - that’s us!) 6
7. Firefox 4 - Motivation • What are those people trying to do? • Users: interact, create, find, learn, experience, control • Developers: build, create, express, impress, reach How can we lead How can Firefox be the Open Web? the best tool for this? What does Firefox need to succeed? 7
8. Firefox Themes for users... for web developers... Fast, friendly and empowering Tools for capable, fast Web apps Simpler, cleaner, faster navigation New HTML5 / Web technologies Control your relationship with websites Native multimedia capabilities Personalize and customize Great developer tools a technology base that’s... Fast, secure, optimized Faster JS, DOM, and UI Stable & secure by design 8 Optimized for today’s hardware
9. Firefox 4 - Plan for Users Fast, friendly and empowering Simpler, cleaner, faster navigation New, sleek, simpler default theme Control your relationship with websites Fewer user interface controls Fewer pixels between user and content Personalize and customize Performance optimizations (user experience) User experience improvements No more modal dialogs No interruptions at startup Updates apply in the background Faster navigation for today’s web users Switch to tab Dedicated “application tabs” PLANS Tab Candy MIGHT CHANGE (please don’t overreport)
10. PLANS MIGHT Simpler, cleaner, faster navigation New sleek, simpler default theme CHANGE (please don’t overreport)
11. PLANS MIGHT Simpler, cleaner, faster navigation Performance optimizations CHANGE (please don’t overreport)
12. PLANS MIGHT Simpler, cleaner, faster navigation Faster navigation for today’s web users CHANGE (please don’t overreport)
13. Firefox 4 - Plan for Users Fast, friendly and empowering Simpler, cleaner, faster navigation Control your relationship with websites Simple, powerful control over websites’ permissions Personalize and customize One-click overview of user/site relationship Backup and share user data with Firefox Sync Single click sign on to websites PLANS MIGHT CHANGE (please don’t overreport)
14. PLANS Managing your relationship MIGHT Control over websites’ permissions CHANGE with websites (please don’t overreport)
15. Firefox 4 - Plan for Users Fast, friendly and empowering Simpler, cleaner, faster navigation Control your relationship with websites Personalize and customize New customization management interface Easier to discover and find relevant add-ons Install add-ons without restart Jetpack based add-ons running out of process PLANS MIGHT CHANGE (please don’t overreport)
16. PLANS MIGHT Personalize and customize New Add-ons Manager CHANGE (please don’t overreport)
17. Firefox 4 for Users: Risks and gaps Theme Current Projects Gaps & Risks • Started theme changes on trunk Simple, clean • Switch-to-tab on trunk, “App Tabs” work started • Work begun to remove modal dialogs • Need to prioritize and resource user experience performance issues • Need to identify measurements & targets for performance issues fast navigation • Work planned for updater improvements • Need to address performance problems from dirty profiles • User experience performance issues identified • Must quickly finalize use cases and designs for permissions manager Giving users • Designsfor user control over permissions started • Weave planning to be ready for first beta • Must resource permissions manager work control & ownership • Account Manager evaluation version released • Solid Weave integration plan • Account Manager specification still in flux Full ability for • New extension manager landed • Jetpack work underway • Compatibility issues from API changes customization • New AMO interactiveness plans begun • Evangelism of Jetpack and API changes PLANS MIGHT CHANGE (please don’t overreport)
18. Firefox 4 - Plan for Web Developers Tools for capable, fast Web apps Bidirectionally connected apps (Websockets) New HTML5 / Web technologies Better AJAX-y interactions (PushState) Native multimedia capabilities New interactions (gesture & multi-touch) Easier layout and styling (CSS3) Great developer tools HTML5 Parser (HTML5) Content creation (contentEditable, HTML5 forms) Structured data storage (IndexedDB) PLANS MIGHT CHANGE (please don’t overreport)
19. Firefox 4 - Plan for Web Developers Tools for capable, fast Web apps New HTML5 / Web technologies Animation of web content (CSS Transitions,SMIL) Native multimedia capabilities Animation API (RefreshDriver) Great developer tools High quality native video playback (<video>) High quality native audio playback (<audio>) Faster 2D drawing (<canvas>) Fullscreen API 3D capabilities (WebGL) PLANS MIGHT CHANGE (please don’t overreport)
20. Firefox 4 - Plan for Web Developers Tools for capable, fast Web apps New HTML5 / Web technologies Native multimedia capabilities Firebug compatibility Great developer tools Remote JavaScript Debugger Web Console Web Inspector New profile manager Timer API Memory Diagnostic Tools PLANS MIGHT CHANGE (please don’t overreport)
21. PLANS MIGHT Great developer tools Console, Inspector CHANGE (please don’t overreport)
22. Firefox 4 for Web Developers: Risks and gaps Theme Current Projects Gaps & Risks • CSS3, HTML5 implementation (partial) • WebSockets, CSP, PushState implementation mostly complete • Need to prioritize and evangelize based on web developer need New HTML5/Web • Multitouch Gestures in DOM underway • IndexedDB specification unlikely to be finalized Technologies • IndexedDB spec & implementation • contentEditable bugfixes, HTML5 forms project specified • contentEditable / HTML5 form work likely to be partial • decisions on shipping implementations of unfinished specifications • CSS3, SMIL and RefreshDriver for controlling animation • OurCSS syntax maps to W3C, doesn’t match WebKit Native multimedia • updated native video support • Fullscreen API specified • videocodec path under investigation capabilities • WebGL implementation started, resourced • WebGL underresourced, driver requirements risky • some user control issues with fullscreen API • Web Console and Web Inspector work started • Inspector is behind schedule Great developer • about:memory partially implemented • Console is not as visually appealing as WebKit tools tools • Firebug compatibility tracking • about:memory needs additional resources PLANS • JSD2 planned, work starting • Profile Manager is unowned MIGHT CHANGE (please don’t overreport)
23. Firefox 4: Plan for Platform JägerMonkey HTML5 Parser off main thread 64 bit support Startup timeline optimizations Reduced I/O operations on main thread JS threads and GC DOM Performance improvements Layers for compositing, scrolling Pathological performance investigations Fast, secure, optimized Faster JS, DOM, and UI PLANS MIGHT Stable & secure by design CHANGE (please don’t overreport) Optimized for today’s hardware
24. Firefox 4: Plan for Platform CSP (allowing secure pages) :visited fix (privacy) ForceTLS (authenticated login) CritSmash Early Security Reviews PAKE (authenticated login) Fast, secure, optimized Faster JS, DOM, and UI PLANS MIGHT Stable & secure by design CHANGE (please don’t overreport) Optimized for today’s hardware
25. Firefox 4: Plan for Platform Graphics compositing with Layers Hardware acceleration using Direct2D Multitouch support Aero Peek integration OSX integration Standards for contacts, cameras, microphones Fast, secure, optimized Faster JS, DOM, and UI PLANS MIGHT Stable & secure by design CHANGE (please don’t overreport) Optimized for today’s hardware
26. Firefox 4 Platform: Risks and gaps Theme Current Projects Gaps & Risks • JägerMonkey work started • JägerMonkey targets still under evaluation Faster JS, DOM and • DOM performance improvements begun • JS threads and GC plan in place • DOM performance measurements hard to prove UI responsivness • Over 50% of I/O off the main thread • Our use of cache is potentially holding us back immensely • Several ideas for responsiveness, but all unowned • Layers for graphic compositing and scrolling improvements Stable and secure • :visited • CSP, change complete ForceTLS implementations well underway • No sandboxing beyond OOPP for Jetpack and plugins by design • CritSmash program tracking to targets • Direct 2D optimizations Optimized for • Harfbuzz for text rendering • Decisions need to be made for supported/unsupported OSes • No plan for Firefox on touch-based devices (Fennec only?) today’s hardware • Multitouch support • Need plan for driver hell in Direct2D space PLANS • Windows 7 feature integration MIGHT CHANGE (please don’t overreport)
27. Firefox 4: Release Milestones 3.6 3.6.1 3.6.3 3.6.4 3.6.5 3.6.6 3.6.7 3.6.8 Firefox 3.6 OOPP mozilla 2010 summit String, UI freeze Firefox 4 frequent beta updates ship A1 A2 A3 A4 A5 Beta RC1 API freeze Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec PLANS MIGHT CHANGE (please don’t overreport)
28. Firefox 4 - Beta Program • Aimed at early adopters and mainstream users • Frequent updates (14 - 21 days) • Interactive • Focused on gathering feedback and data from “the field” PLANS MIGHT CHANGE (please don’t overreport)
29. Firefox 4 - First Beta (OMGNOTFINALDONOTQUOTE ) just beltzner’s proposal • Most of the new theme for user evaluation • New extension manager for user / add-on developer evaluation • Tab management improvements, basic application tab support • HTML5 parser, CSS3 transitions, CSP, WebSockets, :visited • “Reference” Layers implementation PLANS MIGHT CHANGE (please don’t overreport)
30. Firefox Platform: Doing two things at once • in 2009 and early 2010 we proved that we can split our focus • once we branch for beta, must have plans in place • priority will be given to full content/chrome process separation • will start developing & communicating those plans next month PLANS MIGHT CHANGE (please don’t overreport)

Web 2.0

Web 2.0

 

1. Web 1.0 was Commerce Web 2.0 is People - Ross Mayfield Web 2.0 seems to be like Pink Floyd lyrics: It can mean different things to different people, depending upon the your state of mind. - Kevin Maney
2. Introduction
   • The concept of &quot;Web 2.0&quot; began with a conference brainstorming session between O'Reilly and MediaLive International in 2004
   • The phrase &quot;Web 2.0&quot; hints at an improved form of the World Wide Web
   • Emphasizing tools and platforms that enable the user to Tag, Blog, Comment, Modify, Augment, Rank, etc.
   • The more explicit synonym of &quot;Participatory Web&quot;
3. Principles of Web 2.0
4. No Products but Services
   • “ There are no products, only solutions”
   • Not what customer wants but why they want
   • A problem solving approach
   • Simple Solutions
5. Customization
   • Every individual is unique
   • Some people want to be different
   • Allow him to choose instead of forcing him to use what you have made
   • Make him feel home
   • e.g.
   • My yahoo, Google Homepage, myspace
   • Firefox extensions
   • e.g. (This slide is better for reading online )
6. Customization
   • Every individual is unique
   • Allow him to choose instead of forcing him to use what you have made
   • Some people want to be different
   • Make him feel home
   • e.g.
   • My yahoo, Google Homepage, myspace, my naukri??
   • Firefox extensions
   • e.g (This slide is better for taking printouts) .
7. Focus on the “Long Tail”
   • Reach out to the entire web
   • To the edges and not just to the centre, to the long tail and not the just the head
   • Leverage customer-self service
   • e.g. Google, StumbleUpon, orkut
8. Harnessing Collective Intelligence
   • Network effects from user contribution are the key to market dominance in Web 2.0 era
   • The Wisdom of crowds – Users add value
   • Amazon, ebay - User reviews, similar items, most popular,
   • Wikipedia – content can be added/edited by any web user,
   • Flickr – tagging images
   • Cloudmark – Spam emails
9. Harnessing Collective Intelligence..
   • Some systems,designed to encourage participation
   • Pay for people to do it – ‘gimme five’
   • Get volunteers to perform the same task
   • Inspired by the open source community
   • Mutual benefits e.g. P2P sharing
10. Harnessing Collective Intelligence
    • But only a small percentage of users will go to the trouble of adding value to your applications via explicit means.
    • Therefore web 2.0 companies set inclusive defaults for aggregating user data and building value as side effect of ordinary use of the application.
    • It requires radical experiment in trust
    • “ with enough eyeballs, all bugs are shallow”
    • - Eric Raymond
11. Specialized Database
    • Every significant application to date has been backed by a specialized database
    • E.g. Amazon, Google, Ebay
    • Database management is the core competency of Web 2.0 companies
    • “ infoware” rather than merely “software”
12. Who owns the data
    • Control over data has led to market control and oversized financial returns
    • It will provide a sustainable competitive advantage to the company
    • Especially is data sources are expensive to create or amenable to increasing returns via network effects
    • Race is to own certain classes of core data e.g. naukri.com, 99acre, yahoo
13. End of the Software Release Cycle
    • “ Release Early and Release Often”
    • “ Perpetual BETA”
    • Daily operations must become a core competency
    • Software will cease to perform unless it is maintained on a daily basis
14. End of the Software Release Cycle..
    • Automate the maintenance process
    • Real time monitoring of user behavior
    • Microsoft – upgrades every 2-3 yr
    • Flickr- Deploy new build up to every half hr
    • “ Put two or three new features on some part of the site everyday, and if user don’t adopt them, take them out. If they like them roll them out on entire site” - Anonymous
15. Software above the level of a Single Device
    • The PC is no longer the only access device for internet applications
    • Applications that are limited to a single device are less valuable than those that are connected.
    • Design your application from the get-go to integrate services across handheld devices, PCs, and internet servers.
16. Benefits of Web 2.0
    • Recruitment:
    • Due to the cutting-edge underlying technologies and usability-focused interfaces (the ‘cool’ factor)
    • Organisations adopting Web 2.0 tend to attract sophisticated, high-caliber technical candidates.
    • Reduced cost:
    • Not only are Web 2.0 offerings low-cost, but the same techniques can also be applied to existing (non-Web 2.0) products and services, lowering costs.
    • For example, wikis can enable your users to build documentation and knowledge base systems, with relatively little investment from yourself.
17. • Loyalty
    • The open, participatory Web 2.0 environment encourages user contribution, enhancing customer loyalty and lifespan.
    • Marketing/PR.
    • By taking advantage of the aforementioned benefits, marketing and PR teams can implement low-cost, wide-coverage, viral strategies.
    • Search Engine Optimisation (SEO)
    • Web 2.0 delivery mechanisms - such as Blogs and RSS - significantly enhance search engine exposure through their distributed nature
    Benefits of Web 2.0
18. Summing Up
    • No products but solutions
    • Customization ability
    • Focus on long tail
    • Users add value
    • Specialized Database
    • Perpetual Beta
    • Software above the level of single device
19. So Far So Good
    • In the year and a half since, the term &quot;Web 2.0&quot; has clearly taken hold, with more than 9.5 million citations in Google.
    2005 2006
20. What Next ??
21. Web 3.0
    • This term that has been coined to describe the Semantic Web
    • It promises to “organize the world’s information”
    • Can reason about information and make new conclusions

Testing Mobile JavaScript

 

Testing Mobile JavaScript

1. Testing Mobile JavaScript John Resig June 2010
2. I’ve been researching and trying to make sense of the mobile space recently. Started preparations in 2009 (with the creation of TestSwarm).
3. Cross-browser mobile web development is crazy. (...but not nearly as crazy as it use to be.) I’ve seen things...
4. A Simple Goal ✤ Started with a simple goal: ✤ To make sure that jQuery works on the most popular mobile platforms and browsers. ✤ As it turns out, it’s really hard to define the scope of the problem.
5. Questions: Answered ✤ Three questions that need to be answered before doing mobile development: ✤ What platforms and browsers are popular? ✤ What browsers are capable of supporting modern scripting? ✤ What devices and simulators do I acquire to test with? ✤ To answer those questions we need good data.
6. Best Stats? ✤ Who has the best statistics on the mobile market? ✤ StatCounter and Gartner seem to be tops. ✤ StatCounter covers billions of hits per month. ✤ Gartner is well-respected and knows sale information well. ✤ AdMob is decent for specific platforms (iPhone, Android). ✤ That being said: Very little actually-useful information leaks out.
7. Lack of Information ✤ Right now finding this information is a challenge. ✤ As a result, developers develop for what’s in front of them:
8. Platforms
9. Platform Sales
10. Hard Question ✤ What versions of those platforms are popular? ✤ No one seems to know, or isn’t talking about it.
11. Browsers
12. Hard Question ✤ What versions of those browsers are popular? ✤ No one seems to know, or isn’t talking about it. ✤ Supposedly Yahoo is going to be releasing some information soon, we’ll see. ✤ Right now it’s considered to be a competitive advantage to keep it private.
13. Testing Strategy ✤ Draw a line in the sand ✤ Buy Devices, Download Simulators ✤ Automate testing (TestSwarm!)
14. Drawing a Line ✤ Yahoo!s Graded Browser Support technique is optimal. ✤ Qualify which browsers you will support (actively test against) and give them a grade. ✤ A = Full support, C = Fall back to old site, Other = Assume full support.
15. My God... it’s full of browsers... Mobile Graded Browser Support: Market share + Browser quality
16. Testing Browsers ✤ Two options: ✤ Buy devices and/or ✤ Download simulators ✤ Simulators are good for most automated testing. ✤ Any sort of interaction testing you’ll want to do on a physical device. ✤ Always good to do a sanity check on a physical device before going live.
17. Simulators ✤ Simulators are available for most platforms and browsers. ✤ Most simulators require Windows to run (some require Perl or Java) ✤ Some browsers even provide standalone executables (Opera, Fennec). ✤ Getting the simulators running can be a real bear.
18. Automated Testing ✤ Once you have simulators (or physical devices) up and running you’ll want to interact with them as little as possible. ✤ Automated test execution will be really important. ✤ TestSwarm was developed for this express purpose: Make it easy to push tests out to a large number of clients (even mobile).
19. Platforms ✤ Symbian (S60 / UIQ) ✤ iPhone OS ✤ Blackberry OS ✤ Android ✤ Windows Mobile ✤ WebOS ✤ Maemo / Meego
20. Symbian
21. Symbian S60 ✤ The most popular mobile OS - heavily used by Nokia. ✤ v5.0 is for touch screen devices (Equiv. to Safari 3.1) ✤ v3.0 is the latest for ‘regular’ devices (Equiv. to Safari 2.0) ✤ 2 ‘feature packs’ have been released ✤ The oldest, active, mobile browser that should be supported. ✤ Simulators available on Nokia.com.
22. Symbian S60 B A B?
23. Symbian UIQ ✤ A now-dead implementation of the Symbian OS last released in 2008. ✤ Appears to still be in use, some what, but quickly becoming irrelevant. ✤ Uses Opera Mobile (8.6, 8.65) as its built-in browser. ✤ UIQ is dead and Nokia doesn’t distribute the simulator any more. I found a guy in Russia that has some extra copies (totally legit, heh).
24. Symbian UIQ C C
25. iPhone OS ✤ Apple’s super-popular mobile OS ✤ Apple is very good about pushing updates, almost all users are on the latest OS version. ✤ Current with Safari 4 ✤ A gotchya: Does not have fixed position support (making it difficult to implement toolbars). Recommend: TouchScroll library. ✤ Simulator is part of the iPhone SDK.
26. iPhone Simulator No 2.x simulator! B A A
27. Blackberry OS ✤ Blackberry continues to be enormously popular - and growing rapidly. ✤ Ships with a custom browser, will be switching to a WebKit-based one in 6.0. ✤ Anything older than Blackberry 4.6 is really frightening. ✤ Andrew Dupont calls 4.6 the “rubicon”. ✤ 4.6/4.7 have some strange quirks, 5.0 is pretty decent. ✤ Blackberry provides simulator downloads on their dev site.
28. Blackberry OS B B B
29. Android ✤ Rapidly growing OS from Google. ✤ Tons of manufacturers are modifying and shipping it. ✤ Expect Android to grow drastically over the next couple years. ✤ Lack of control has yielded extreme version fracturing. ✤ Simulators are an easy download.
30. Android A A A
31. Windows Mobile ✤ Windows Mobile 6.5 is the current release. ✤ Uses IE 6 as its rendering engine. ✤ Window Mobile 7.0 is coming soon. ✤ Uses IE 7 as its rendering engine. :-( :-( :-( :,( ✤ Window Mobile 6.1 is still relatively popular but is based off of IE 4.0 - this is a non-starter. ✤ Windows Mobile 6.5 simulator is an easy download.
32. Windows Mobile 7.0? :-( F B
33. WebOS ✤ Palm’s (HP’s?) Mobile OS ✤ Has a minimal amount of traction. ✤ Built off of WebKit, easy to support. ✤ 1.4 simulator ships with 1.3 browser (wtf!?) ✤ I’m testing using a physical Palm Pre. ✤ Palm Pre simulator is an easy download. A
34. Maemo / Meego ✤ Linux-based OS used for tablets (and potentially phones). Used by Nokia. ✤ Not very popular. ✤ Has a browser called ‘MicroB’ that C uses Gecko. ✤ Mostly interesting as Fennec is capable of running on it. ✤ Simulator requires Linux to run - haven’t figured it out yet.
35. Platform-Independent Browsers ✤ A number of browsers target multiple platforms. Some work directly with carriers (Opera, Netfront) whereas others provide downloads (Opera, Fennec). ✤ Opera ✤ Fennec ✤ Netfront ✤ Phonegap (App Platform)
36. Opera Mobile and Opera Mini ✤ Hugely popular mobile browser (especially Mini). ✤ Available on a large number of platforms. ✤ Even ships as the default browser on some. ✤ Opera Mobile and Opera Mini couldn’t be more different. ✤ Mobile is a full-featured browser. ✤ Mini is a glorified bitmap viewer. ✤ Pulls from a proxy, no JS executing on client.
37. Opera Mobile and Opera Mini C A C Simulator is awesome!
38. Fennec ✤ Codename for Firefox on mobile devices ✤ Released for Maemo, alpha release out for Android (watch this!) A B ✤ Equivalent to the latest releases of Firefox Simulator is awesome!
39. Netfront ✤ Used on Playstation and N-Gage ✤ Download available for Windows Mobile. ✤ Highly crippled, custom, browser. ✤ Not worth supporting. C ✤ Compatibility can be a real pain.
40. Phonegap ✤ Not a browser, a way to develop cross-platform applications. ✤ Uses WebKit as the rendering engine. ✤ Use HTML/CSS/JS to develop deployable apps. A ✤ Quite popular, easy to support.
41. Other Browsers ✤ Not worth the time for such small market share. ✤ Ozone (Uses WebKit) ✤ Openwave (Dead) ✤ Bolt (Uses WebKit) C ✤ JIL (Uses WebKit, Widget Platform)
42. Summary ✤ What do we need to support? Roughly: ✤ Opera 9.5, 10 ✤ Internet Explorer 6, 7 ✤ Firefox 3.5+ ✤ Safari 2, 3, 4 ✤ (and Blackberry... weeee...) ✤ Don’t assume it’ll work, download the simulators and test!
43. What Devices to Buy? ✤ For A level: ✤ 1 iPhone 3GS: Latest running 3.x (* Watch for new 4.x iPhone) ✤ 1 iPad: Running 3.2 ✤ 1 Nokia N97 Running S60v5 ✤ 1 Palm Pre: Running 1.4 ✤ 2 Android Phones: HTC Magic (1.5/1.6), Droid Incredible (2.1) (* Watch for new 2.2 Android) ✤ And download the Opera and Fennec Simulators and Phonegap
44. A -Grade Costs ✤ $315 - iPhone 3GS ✤ $499 - iPad ✤ $370 - Nokia N97 ✤ $399 - Palm Pre ✤ $349 - HTC Magic ✤ $637 - Droid Incredible ✤ Total: $2569
45. What Devices to Buy? ✤ For B level (in addition to A level): ✤ 1 1st Gen iPhone: Running 2.x ✤ 1 Nokia N96: Running 3.0v2 ✤ 3 Blackberry Devices: Curve 8900 (4.6), 2x Tour 9630 (4.7 and 5.0) ✤ 1 HTC HD2: Running Windows Mobile 6.5

Know All about Google

All about Google

 

1. Everything you always wanted to know about Google…But were afraid to ask Paris, December 2008
2. This work is licensed under the Creative Commons to allow for further contributions by other specialists and web users in the coming months. To view a copy of this Attribution-NonCommercial- ShareAlike3.0 Unported license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/ or send a letter to Creative Commons, 1712nd Street, Suite300, San Francisco, California, 94105, USA. ..……. December 2008 • Everything you always wanted to know about Google… • 2
3. Google key success factors : Web specific ? Scalability Network effects Data mining • Ability to easily grow at • The utility of a good or a • The web offers the marginal costs service varies with the opportunity to exploit and • Applied to infrastructures : number of users analyze a very large ability to adapt its size to • The reach of a critical amount of data high load & volumes mass of users constitutes • Users’ behavior can be • Applied to business a significant barrier to the analyzed to create models : ability to entry monetizing value monetize millions of users Openness Cocreation Business model • The traditional walled • Non-traditional actors • Advertising is not a market garden1 media strategy become part of the value but a business model becomes irrelevant chain • Any market that attract • Content and services must • Users, content creators advertising is a target for be open and interoperable and external developers Google to favor audience are given the tools to circulation create new markets and enrich services ..……. 1 Network or portal which offers only its own content or services to users December 2008 • Everything you always wanted to know about Google… • 3
4. 1 Why won’t Google be affected by the crisis ? 2 Why is Google trying to change the mobile world ? 3 Why isn’t YouTube a content portal ? 4 Why does Microsoft fear Google ? 5 How Google wants to compete with Facebook ? 6 Why is Google buying satellites ? 7 How does Google buy traffic ? 8 Why did Google acquire DoubleClick ? 9 Why doesn’t Google monetize all its services ? 10 How does Google capitalize on Open Source developers work ? 11 How did Google capture the offline advertising market ? 12 Why is Google stealing our voices ? 13 Is PageRank a really competitive advantage ? 14 How does Google turn advertising into information and performance ? Annex: Network effect, two-sided market, glossary, financial , contact ..……. December 2008 • Everything you always wanted to know about Google… • 4
5. 1 Why won’t Google be affected by the crisis ? 4 levers will allow Google to increase its revenues amidst the economic crisis Performance Bn$ 35 1 advertising Licenses and During a crisis, other performance revenues advertising gains market shares Internet Explosion of the Mobile non advertising + 3% 2007-2008 1 revenues On line Monetisation of +450% 2007-2008 video the Internet Mobile audience YouTube Bn$ 201 monetization 260 millions mobile subscribers increases worlwide oct 2008 1 Estimates 2008 : Revenues = 200 M $ 1 Revenues 2008 Revenues 2012 Google is in a situation in which it can resist the economic crisis and find new revenue sources, both advertising and non-advertising ..……. 1: faberNovel estimates December 2008 • Everything you always wanted to know about Google… • 5
6. 2 Why is Google trying to change the mobile world ? (1/2) Mobile industry is based on a “ traditional” locked values chain. Operating Terminal Portal Operator Web Contents System Examples: Entry barriers stop Google from applying its models to the mobile industry: Access restricted to services/contents Limited Network Access and to their improvement • Terminals assigned to a unique operator • Operator portals favored over other portals • Difficulty of interconnecting networks • Services offered by terminal manufacturers favored over other services • Closed operating systems The mobile industry is not suited for the Google development model based on openness, interoperability and network effects. ..……. December 2008 • Everything you always wanted to know about Google… • 6
7. 2 Why is Google trying to change the mobile world ? (1/2) In order to duplicate its open model, Google designed a three fold strategy : Android : Open Source Lobbying Telco partnerships O.S.1 • Android Developer Challenge: • Google candidacy for mobile • Pressure from Google to force Contest for developers to create license attribution aims to force the operators to offer its applications new applications for Android FCC2 to impose an openness as default options clause to the winner • Open Handset Alliance: Common initiative of 34 mobile • Sharing of advertising phone industry players • A success : clause partially revenues between Google and (manufacturers, suppliers and imposed on the winning bidder, operators distributors) aiming at spreading Verizon. Android In addition, Google developed and acquired mobile devices applications : Location-based Social Collaborative Games services Networks Tools ShareYour Grand central 3 Cab4me FreeFamilywatch Golfplay JOYit Wertago Jaiku 3 Board Google breaks open the mobile industry value chain to create an environment that will be fit to the distribution of its products and monetization model. ..……. 1 Operating System 2 Federal Communications Commission 3 Acquisitions December 2008 • Everything you always wanted to know about Google… • 7
8. 3 Why isn’t YouTube a content portal ? (1/2) YouTube acquisition is part of a strategy to monitor key content and audience hubs. Monetise contents Give access to Organise information through various information sources of audience •A new information silo: video • YouTube bandwidth • Broadcasting of contents search spending estimated to reach through Google websites as 1M$/day1 well as other sites •Next : Speech to text technology : information • Revenue sharing logic searches within video contents YouTube has already won the audience battle … 75% … 60% May 2008 16% 9% 4%8% 2% 3% 1% 1% May 2007 Market shares of 5 leading video websites, United States (may 2008 vs.may 2007) [%]2 1 NewYork Times ..……. 2 Hitwise 2008 December 2008 • Everything you always wanted to know about Google… • 8
9. 3 Why isn’t YouTube a content portal ? (2/2) Unlike its competition, YouTube follows an open logic and focuses primarily on developing content distribution tools : Encourage content Attract as many viewers as Monetize through relevant providers to use the service possible advertising tools • YouTube Program Partner : • Broadcasting videos on • Traditional advertising: Payed providers of semi- YouTube website and other AdSense and banners professional content Google sites (ex : Google.com, GoogleNews) • In-video advertising : Pre-roll, post-roll, overlay2 • Broadcasting contracts • Exporting videos (blogs, with major content providers social networks) and • Brand advertising: developing API1 for broadcasting video ads advanced broadcasting on within an environment third party websites coherent with the brand’s image • Broadcasting on all video devices : television, mobiles, • E-Commerce: Affiliation of multimedia players, video partner websites (Amazon, consoles Itunes, video games) YouTube acts as the platform of a two sided3 market composed of content providers and video seeking users. 1 Application Programming Interface. Standardized programming protocol allowing applications to communicate 2 Clickable text advertising displayed on a video 3 See Annex ..……. December 2008 • Everything you always wanted to know about Google… • 9
10. 4 Why does Microsoft fear Google ? Google’s ambition isn’t limited to “in browser” Web services, but extends to any online or offline application market. Google sets out to enter the online application market, MSFT’s cash cow : Disruption of offline application Consolidation of the online market segments environment • Launching of the Office Google • Launching of Google Gears: software pack: a word-processor, Open Source project allowing an a spreadsheet program, a offline use of online applications presentation tool and a calendar • Free alternative to Microsoft Office pack • Claimed ambition of becoming a standard and encouraging online languages as opposed to offline programming languages • Acquisition of SketchUp, a 3D modeling software with a free version made available Google partly « destroys » Microsoft’s market when shifting value from offline to online ..……. December 2008 • Everything you always wanted to know about Google… • 10
11. 5 How Google wants to compete with Facebook ? In 2007, Google launched Opensocial : a series of multi platform API allowing developers to create compatible applications with partnering social networks, Ex : Slideshare application available on Linkedin and hi5 OpenSocial is not a Facebook competitor but a «meta-social network ». Network Network effects effects Network effects Google wants to become the « social data search engine » and to monetize this data, leveraging network effects. Facebook’s platform is limited, Google’s is the whole Web ..……. December 2008 • Everything you always wanted to know about Google… • 11
12. 6 Why is Google buying satellites ? (1/2) Since 2005, Google has been multiplying investments in all kinds of infrastructures: • Free Wifi in Mountain View • Investment in Fon: shared Wifi access Wifi Gratuit Satellite Wimax • 60 M$ Investment • Partnership • Internet access in • A mobile high-speed developing countries internet access technology • A 100 M users Baloon market by the end of Backbone1 Wifi 2008 • Partnership • 100 M$ investment announcement • Expansion of high- • Internet access speed networks technology superior to satellites for isolated areas ..……. 1 Long distance high-speed networks, core of the Internet network December 2008 • Everything you always wanted to know about Google… • 12
13. 6 Why is Google buying satellites ? (2/2) Internet infrastructures is actually Google’s business infrastructure: Traffic on Google websites depends on internet infrastructure development and availability (Backbone, Wifi, Satellite,…) Google has three objectives when investing in the upstream part of its value chain: Strengthen and secure existing • Strengthening and securing existing infrastructure infrastructure lightens Google’s dependancy on its providers • Favouring high-speed Internet access means more Favor high-speed Internet access time spent and usage volumes, thus increases Google services usage Prioritize Internet access for • Future web users are Google services’ next unconnected countries or populations users Through infrastructures investments, Google reinforces its traffic providers and increases access to its services. ..……. December 2008 • Everything you always wanted to know about Google… • 13
14. 7 How does Google buy traffic ? Google asserts its ability to attract users on the sole basis of its services quality and without turning to advertising… Truth is Google largely buys traffic from providers Browsers Manufacturers Toolbar Portals •Firefox: 60 million daily • In 2006, partnership users in 2008 deal with Dell to have the • The Google Toolbar is part of the web navigator, •Google finances 85% of Google search engine which makes Google the • In 2005, Google bought Firefox in exchange for appear by default on Dell a 5% stake in AOL for 1 computers default search engine having its search engine billion$ (20 million embedded in the browser • In 2008, partnership • Adobe installs it as part subscribers at time of deal) of a package with deal with Apple to have the • Google became AOL’s Google search engine Shockwave(2006) white label search engine appear by default on • Sun has been installing • Google expands its Iphones (13 million devices it as part of a package with advertising network reach sold by october 2008) Java since 2005 (20 million • Partnerships with uploads/month) manufacturers allow the search engine to be guaranteed to in a prime position. Google has the financial power to buy traffic from partners, accessing to massive audiences. ..……. December 2008 • Everything you always wanted to know about Google… • 14
15. 8 Why did Google acquire DoubleClick ? • In 2007, Google acquired DoubleClick, one of the world leaders in display advertising : To position itself on the banner To reach for highly popular market websites Google « moves up » the long tail Market shares of main online of advertisers advertisers in relation to website traffic 1 Ad budget per advertiser Advertising < 100k 100k- >1Mon houses UV2 1M UV2 UV2 Banner Market Adbrite 4,1% 4,9% 0,5% Text ad market AOL 1,9% 6,5% 5,7% DoubleClick 9,1% 29,9% 48,0% Google 71,4% 41,6% 15,8% advertising Traditional MSN 6,6% 6,3% 12,8% market Yahoo 4,7% 7,3% 16,5% Number of advertisers Google acquired DoubleClick to gain an expertise (display) and global market shares (highly popular websites) 1 www.attributor.com/blog/2008/03/ 2 Unique Visitors ..……. December 2008 • Everything you always wanted to know about Google… • 15
16. Why doesn’t Google monetize all of its 9 services? Some Google services are free of charge and monetized through advertising : Blogger: blog creating tool Google Health: service for managing and storing personal medical information GoogleNews: personalized mash-up of news articles and summary Picasa: photo sharing service Google SketchUp: 3D model creating tool Goog-411: phone information service … These services are actually indirectly monetized : Tools designed to Products specifically Attract new customers generate audience are developed to improve through loss leaders made available other Google products Monetization of blogs created on Google 411 created to better the Picasa devised as a loss leader Blogger through AdSense or video indexing on YouTube towards other Google products FeedBurner Google global strategy allows strong indirect monetization of its products ..……. December 2008 • Everything you always wanted to know about Google… • 16
17. 10 How does Google capitalize on Open Source developers work ? Google encourages development of open source applications: • Google code: platform designed for Open Source developers Supplied with guides, tutorials, code extracts of Google products • Google Search Code: code search engine Automatic referencing of all code sections that can be found on the Internet • Events created for the Open Source community: Google Summer of Code: grants awarded to Open Source student projects Google Developer Days: seminars dedicated to Google products Through support of Open Source community, Google pursues 4 objectives: Development of Open Increase of total Promotion of a more Assembling a free Source langages used Internet traffic « open » Web public relations team by Google •New applications •Increasing the •Developers’ chats are •The number of create new uses, interoperability a very effective public available Open Source leading to increased multiplies network relation tool codes encourages the total traffic effects1 emerging of new •Opensource is products becoming an •Feedback of Open advantage to attack Source developers proprietary code helps creating new strategies products Google supports the Open Source community in a spirit of collaborative creation, one of Google’s strategic pillars ..……. 1 See Annex December 2008 • Everything you always wanted to know about Google… • 17
18. 11 How did Google capture the offline advertising market ? Online advertising market accounts for only 8% of the US market1: 3% 7% 8% Billboards Radio Most offline media (television, 44% 17% Internet Daily Press radio, press,…) begin to be IP- News Press Television ready with online versions 21% Google is exploring the offline ad market! • Ad transfer from online to offline (YouTube on television) • Entry on traditional offline markets (radio, billboards,…) • Integration of offline techniques (traditional fixed pricing) • Partial adaptation of AdWords onto radio and television Google’s entry on this market anticipates new uses and broadens its offer. Not specifically successful for now… ..……. 1 TNS Media Intelligence, US advertising market December 2008 • Everything you always wanted to know about Google… • 18
19. 12 Why is Google stealing our voices? In 2007, Google launched Google Voice Local Search in the United States, a free and automated phone directory service : What is the business model of this free and ad-free service ? Google is creating a database of phonemes, recorded during calls in order to better its speech to text1 technologies: Creation of a Development Indexing of Indexing of all of « speech to phoneme text » YouTube audio/voice database audio tracks sources technologies 2 Bla bla bla Externalizing tasks onto users (« crowdsourcing 3») is a commonly used process by Google to improve its products. 1 Converting oral information into text 2 First experiments with political videos posted during the presidential campaign of 2008. 3 Neologism created in 2006 by Jeff Howe and Mark Robinson, Wired magazine editors ..……. December 2008 • Everything you always wanted to know about Google… • 19
20. 13 Is PageRank a really competitive advantage ? (1/2) PageRank is Google’s link analysis algorithm that measures the probability that a page will be relevant to user’s query : Based on the correlation between the amount of links towards a page and their relevance It accounts for the notoriety of the sites that link to the page in question PageRank’s simplified formula is : (A page’s (u) PR is the sum of all PRs of pages linking to u (v), divided by the respective number of outbound links contained in pages v) Google1 claims that PageRank is one of its search engine’s main competitive advantage : A « champion of The search engine’s A tool unlikely to be democracy » « cornerstone » tampered with 1 http://www.google.fr/why_use.html ..……. December 2008 • Everything you always wanted to know about Google… • 20
21. 13 Is PageRank a really competitive advantage ? (2/2) A tool that can’t be tampered with? An outdated technology? Sale or exchange of famous website links Launch of search engines that (webringing) don’t use tools such as Search Engine Optimization1 techniques PageRank (Cuil, Powerset) Google’s search engine success relies on other factors : Relevance guaranteed Scalable architecture Quick/simple queries by 200 other criteria • Capacity of increasing/ • Clear query interface • The search engine’s growing according to the • Simple and quick algorithm was subjected to volume of indexed pages presentation of results 450 modifications in 2007 and number of queries • 2 millions servers by 2008 PageRank is only one of many Google’s search competitive advantages. It is certainly not the main entry barrier to competitors on the search market. 1 Set of techniques aiming at improving a site’s referencing on a search engine. ..……. December 2008 • Everything you always wanted to know about Google… • 21
22. 14 How does Google turn advertising into information and performance?(1/2) « Advertising income often provides an incentive to deliver poor quality search results» Sergey Brin & Larry Page Most advertising models present strong weaknesses Undefined Strong intrusion Weak relevance performance • Large size adverts • Influence results by • Impossible to reward • Slows down results making paid for clients efficient advertising display from a search appear first • Example : television ads engine query • Example: Opentext/Kelkoo • No direct measure of real • Little or non-existent performance targeting • Example : invoicing according to the number of displays (CPM) ..……. December 2008 • Everything you always wanted to know about Google… • 22
23. 14 How does Google turn advertising into information and performance?(2/2) Displays advertising defines Relevance Quality Score Performance For the user Direct impact on advertising value For the advertiser Most relevant ad Qualityscore measures the relevance of the ad and is determined by the click throug rate on the ad. It impacts the display rank and Cost per Click: • Rank: relevant ads are pushed up, non relevant ads do not appear • Cost per click of the ad : performant Total cost of the advertising ads are charged less campaign is determined by the number of clicks on the ads and Less relevant ad not by the number of displays. Google’s advertising model benefits the user (improved relevance) as well as the publisher (performance based billing & rebates) ..……. December 2008 • Everything you always wanted to know about Google… • 23
24. Google : the network circulation value creation model Traditional value creation Network value creation Eg : Microsoft Eg : Google $ Gmail $ Apps Server Products $ Apps $ $ $ Search $ Customer engine Business products Products $ $ $ $ $ $ Partners Entertain Youtube ment $ The global value of the company relies The global value of the company relies on independent lines of on traffic between network parts products/business units (proprietary or partners) ..……. December 2008 • Everything you always wanted to know about Google… • 24
25. Going further : are Google’s key success factors limited to Google • Our conviction : Every company innovating in the digital industry must address and capitalize on the 6 identified key success factors to perform on digital markets. • Our proposition : faberNovel proposes to align companies strategy, development projects and existing products with these factors to ensure success optimization and market performance. Download our comprehensive White Paper « Google’s key success factors » http://www.fabernovel.com ..……. December 2008 • Everything you always wanted to know about Google… • 25
26. Annex • Definition : network effects • Definition : two-sided market • Pricing of a two-sided market • Glossary • Financial datas • Acknowledgement • Contacts ..……. December 2008 • Everything you always wanted to know about Google… • 26
27. What is a network effect? • A network effect describes how a service becomes more valuable to its users as more people use that same service Phone’s utility is limited if The utility for a user According to Metcalfe’s the network is composed raises if the network law, the utility(U) of a of 2 users broadens network is proportional to the square of the number (n) of its users U ≈ k*n2 Network effects creates critical masses of users. They represent significant barriers to entry for competitors. ..……. December 2008 • Everything you always wanted to know about Google… • 27
28. What is a two-sided market? • A two sided market consists of a platform allowing 2 groups of clients/providers to interact and which optimizes the revenue distribution among these groups with the objective of maximizing market sizes. Exemple of the video game market Crossed network effects Internal Price A Price B network Side A Side B effects Developers Consumers Network effects in action Internal effects Crossed effects • A potential market for a • Utility to a new developer developer gets bigger as more increases as the community of consumers enter the said market developers grows (shared knowledge) • The number of games available to the consumer increases with • Utility to a new consumer the number of developers increases as the community of working on the platform consumers grows (secondhand market) ..……. December 2008 • Everything you always wanted to know about Google… • 28
29. Pricing of a two-sided market Price A > 0 Price A < 0 Side A Platform1 Face A Platform1 • The platform may charge a side • … or subsidize a side • Example 1 : Apple Appstore charges • Example 3 : Google Android finances developers by taking a revenue share developers through a contest on sold applications • Example 4 : Youtube finances video • Example 2 : Microsoft makes content providers consumers pay for video games Who should you be charging? Who should be subsidized? The platform must subsidize the groupe that is most price-sensitive and charge the group that is most sensitive to the other group’s size. 1 Provided the sum of price A + price B is a fixed figure, a platform financing a group automatically charges the other group and vice versa ..……. December 2008 • Everything you always wanted to know about Google… • 29
30. Financial data (1/2) Revenues and Net margin (M$) Revenue distribution/activity (B$) 439 1466 3189 6139 10604 16594 Total 16,6 Revenues 0,2 1% CAGR Revenues 02-07 : 107% Net CAGR Net margin 02-07 : 112% Licences and other 5,8 34% margin revenus Partner websites (AdSense) 10,6 4203 3077 Google web sites 1465 65% 100 106 399 CA 2007 2002 2003 2004 2005 2006 2007 Net margin’s growth outperform Most of Google’s revenues still revenues’ growth come from its own websites 78% 79% 85% 2005 2006 2007 Advertising revenue share to partner web sites is increasing2 and is the highest of the market ..……. Source: Google financial tables 2007 2 : calculated as Traffic Acquisition Cost/Google Network web sites revenues December 2008 • Everything you always wanted to know about Google… • 30
31. Financial data (2/2) Advertising revenues Turnover Traffic acquisition costs R&D investments 6139 10604 16594 16,4 10,5 30,1% 599 1229 2120 34,9% 6,1 31,5% 9,0% 11,0% 12,0% 2,1 3,3 4,9 2005 2006 2007 2005 2006 2007 Controlled traffic acquisition costs High R&D investments 34,9% of advertising revenus in 2005 vs 30,1% in 2007 +88%/year between 2005 and 2007 + 26% 1,9 2,4 2005 2006 Data center costs are under control In Billion $ : +26% 2006/2007 (vs +72% revenues) ..……. Source: rapport annuel Google 2007 December 2008 • Everything you always wanted to know about Google… • 31
32. Glossary • API : application programming interface. Standardized programming protocol allowing applications to communicate • Internet Backbone : main trunk connections of the Internet, made up of a large collection of interconnected high-capacity data routes and core routers that carry data across the countries and continents • Crowd sourcing : act of outsourcing a task to users • Data center : facility used to house computer systems and associated components • O.S. : operating system • Overlay ad : clickable text advertising displayed on a video • Scalability: property of a process, which indicates its ability to handle growing amounts of work easily • Search Engine Optimization : process of improving the volume and quality of traffic to a web site from search engines via natural search results (as opposed to paid search results) • Speech To Text : technology converting spoken words to machine-readable input such as text • Walled Garden : closed or exclusive set of information services provided for users by a network or portal ..……. December 2008 • Everything you always wanted to know about Google… • 32
33. Acknowledgements To faberNovel contributors : • Amaury de Buchet, VP Consulting • Cyril Vart, VP Strategy & Development • Alexis Arquié, Junior Project Analyst • Mounir Fassouane, Junior Project Analyst To the bloggers : • Olivier Ertzscheid from affordance.typepad.com/ • Google Operating System : googlesystem.blogspot.com/ • Richard MacManus from readwriteweb.com/ • Techcrunch.com & Mobilecrunch.com/ • Frédéric Cavazza from fredcavazza.net/ ..……. December 2008 • Everything you always wanted to know about Google