This document covers research on how click fraud works, click fraud history and Google strategy towards click fraud.

1.
CLICK
FRAUD
REPORT
EXECUTIVE
SUMMARY
This
document
covers
research
on
how
click
fraud
works,
click
fraud
history
and
Google
strategy
towards
click
fraud.

2. CLICK
FRAUD
REPORT
EXECUTIVE
SUMMARY
Competitors
of
AdSense
publishers
Click
Fraud
usually
requires
a
lot
of
IP’s
(spread
geographically)
to
fool
advertising
companies;
one
of
the
best
ways
to
stop
this
is
by
giving
IPs
to
requesters
with
strong
justification
regarding
utilization.
For
more
details
please
refer
Appendix
A.
These
persons
may
wish
to
Data
centers
are
not
giving
away
IPs
in
excess
of
8
or
10.
This
can
be
exceeded
based
on
personal
or
organizational
relationships.
AdSense
and
other
banner
For
any
additional
IP
request
one
needs
to
submit
form
to
ARIN
(For
USA)
explaining
in
detail
how
we
plan
to
utilize
these
IP’s.
One
needs
to
confirm
that
at-­‐least
80%
to
85%
of
all
IPs
will
be
utilized
and
will
not
be
misused.
For
more
details
please
refer
Appendix
B.
Google
overtime
has
improved
invalid
click
tracking
and
for
some
accounts
it
has
been
observed
that
a
good
number
of
clicks
were
marked
as
invalid.
For
more
details
please
refer
Appendix
C.
Google
uses
sophisticated
algorithms
with
a
high
degree
of
scientific
accuracy
for
protection
against
invalid
clicks.
For
more
details
please
refer
Appendix
D.
frame
online
website
,
which
generate
revenue
through
advertising.
It
is
made
to
look
like
website
(Keyword)
owner
is
clicking
on
its
own
ads
to
make
profit.
The
advertising
network
may
terminate
may
then
terminate
the
relationship
with
that
particular
website.
Click
fraud
is
rising
and
so
are
Google’s
efforts
to
curb
the
nuisance.
For
more
details
please
refer
Appendix
E.
1

3.
HOW
CLICK
FRAUD
WORKS
(APPENDIX
A)
There
is
a
huge
pool
of
marketing
techies
who
have
mastered
this
art
of
click
fraud.
Normally
they
create
a
network
of
users
who
make
profits
by
clicking
on
ads.
There
might
be
hundreds
or
even
thousands
of
people
who
just
do
nothing,
but
click
on
sites.
Google
and
yahoo
have
been
trying
hard
to
cut
down
click
fraud,
but
it
is
virtually
impossible
for
them
to
nullify
click
fraud.
It
is
estimated
that
10%
to
15%
of
all
clicks
are
fake,
and
300-­‐500
million
dollars
of
advertising
budget
get
exhausted
each
year
into
click
fraud.
However,
there
are
various
benchmark
tests
which
you
can
follow
to
detect
fake
clicks.
These
are
as
given
below:-­‐
•
•
•
•
If
you
observe
an
unusual
click
trend
or
CTR
(Click
through
Rate)
in
your
ad
words
account
then
it
could
be
an
indication
of
click
fraud.
You
should
manually
monitor
your
ad
words
account
to
detect
click
fraud.
In
case
you
observe
a
sudden
increase
in
no
of
clicks
within
shot
timeframe
of
15-­‐30
minutes
during
a
day,
then
it
could
an
indication
of
unusual
click
trend.
Moreover
if
none
of
those
clicks
actually
convert
into
lead,
then
it
could
be
an
indication
of
click
fraud.
You
are
receiving
5000-­‐15000
clicks
from
AdSense,
but
none
of
those
clicks
actually
get
converted
into
lead.
Moreover,
if
you
observe
sudden
drop
in
conversion
rates,
then
it
could
be
an
indication
of
click
fraud.
Normally
fake
clicks
originate
from
1
or
2
particular
locations
(depending
upon
your
competitors/industry).If
you
observe
a
regular
pattern
of
several
clicks
from
one
geo
location,
but
none
of
those
actually
converting
into
lead,
then
it
could
be
an
indication
of
click
fraud.
You
should
keep
tabs
on
clicks
from
suspicious
or
untraceable
IP
addresses.
There
are
business
sectors
like
online
marketing
website
development
where
click
fraud
is
really
high.
Click
fraud
is
also
evident
in
other
business
fields
2

4.
ARIN
(APPENDIX
B)
ARIN
provides
services
related
to
the
technical
coordination
and
management
of
Internet
number
resources.
The
nature
of
these
services
is
described
in
ARIN's
mission
statement:
ARIN,
a
nonprofit
member-­‐based
organization,
supports
the
operation
of
the
Internet
through
the
management
of
Internet
number
resources
throughout
its
service
region;
coordinates
the
development
of
policies
by
the
community
for
the
management
of
Internet
Protocol
number
resources;
and
advances
the
Internet
through
informational
outreach.
These
services
are
grouped
in
three
areas:
Registration,
Organization,
and
Policy
Development.
Registration
Services
Registration
Services
pertain
to
the
technical
coordination
and
management
of
Internet
number
resources.
They
include:
•
IP
address
space
allocation,
transfer,
and
record
maintenance
•
ASN
allocation,
transfer,
and
record
maintenance
•
Directories
•
Registration
transaction
information
(WHOIS)
•
Routing
information
(Internet
Routing
Registry)
•
DNS
(Reverse)
•
Who
Was
3

5.
GOOGLE’S
TOP
PRIORITY:
(APPENDIX
C)
The
security
of
Google
Ad
Words
advertisers
is
a
top
priority
for
Google,
and
Click
Fraud
is
a
growing
concern
for
small
business;
with
research
showing
a
rise
of
click
fraud
by
almost
20%
in
2010.
Some
competitors
will
deliberately
generate
and
click
on
your
ads
repeatedly
in
an
attempt
to
exhaust
your
daily
budget
sooner
and
bump
up
the
costs
of
running
your
Ad
Words
account;
this
is
click
fraud.
Certain
patterns
in
your
Ad
Words
account
can
be
observed
to
determine
if
you
are
experiencing
this
type
of
fraud,
for
example,
you
may
notice
a
sudden
and
large
increase
in
Impressions
and
Clicks
for
a
keyword
of
high
cost
and
competition.
There
are
ways
to
ensure
that
a
competitor
isn’t
able
to
perform
this
fraud;
which
you
can
do
yourself.
Google
Ad
Words
account
holders
have
the
option
to
exclude
up
to
100
Internet
Protocol
(IP)
addresses
from
seeing
Ads
within
their
campaigns.
All
Ads
in
the
campaign
are
prevented
from
being
displayed
for
users
with
the
IP
addresses
that
you
specify,
so
choose
your
list
carefully.
An
important
note
about
IP
address
exclusion:
Some
large
internet
service
providers
(ISPs)
use
a
range
of
IP
addresses
for
all
their
users.
If
you
exclude
an
IP
address
that
is
a
proxy
for
many
users,
you
could
be
blocking
a
large
amount
of
legitimate
and
potentially
profitable
traffic.
Google
takes
no
responsibility
for
this
action.
In
saying
all
this,
Google
does
have
its
own
measures
in
place
for
preventing
click
fraud.
Each
click
you
get
is
examined
by
a
system
of
theirs
that
looks
at
the
IP
address,
time
of
click,
any
duplicate
clicks
and
a
few
other
click
patterns.
These
factors
are
then
analyzed
by
their
system
to
isolate
and
filter
out
any
patterns
that
appear
to
be
deliberately
driving
up
the
costs
of
advertisers’
clicks/impressions.
4

6.
DETECTION
OF
INVALID
CLICKS
–
EXTRACT
FROM
GOOGLE
ADWORDS
WEBSITE
(APPENDIX
D)
The
systems
and
processes
that
Google
uses
to
detect
and
filter
invalid
clicks
can
be
divided
into
three
types:
Proactive
1.
Filters
•
•
Automated
algorithms
which
filter
out
invalid
clicks
in
real
time,
before
they
are
charged
to
advertisers’
accounts
Analyze
all
clicks
on
Ad
Words
ads
•
Automatically
discard
clicks
and
impressions
from
known
sources
of
invalid
activity
•
Reactive
2.
Offline
analysis
•
•
•
Automated
algorithms
and
manual
analysis
Focuses
on
the
AdSense
network,
but
also
includes
Google
sites
and
those
in
other
networks
Accounts
for
a
much
smaller
proportion
of
invalid
clicks
than
that
detected
by
our
automated
filters
3.
Investigations
•
Relatively
rare
•
All
advertiser
inquiries
about
invalid
clicks
are
investigated
by
Google
and
taken
very
seriously
•
The
number
of
invalid
clicks
detected
by
investigations
is
negligible
relative
to
the
number
detected
by
filters
and
offline
analysis
Account
for
vast
majority
of
invalid
click
detection
How
Advertisers
are
protected:
The
Ad
Traffic
Quality
Team
helps
protect
advertisers
against
click
fraud
by
working
to
isolate
and
filter
out
potentially
invalid
clicks
before
they
ever
reach
your
account
reports.
Although
the
intent
of
a
click
is
difficult
to
determine
with
a
high
degree
of
scientific
accuracy,
we
try
to
identify
suspicious
clicks
as
often
as
possible
and
to
exclude
additional
clicks
which
we
know
will
perform
poorly
for
advertisers.
By
casting
a
wide
net,
we
try
to
maximize
the
amount
of
invalid
clicks
that
we
exclude,
which
helps
us
to
deliver
consistent
and
high-­‐quality
traffic
to
advertisers.
Undoubtedly,
some
portion
of
the
invalid
click
activity
is
not
detected
by
our
proactive
systems
and
processes.
However,
we
are
able
to
effectively
minimize
that
proportion
through
our
proactive
monitoring
of
both
advertiser
and
publisher
networks.
Our
belief
in
the
quality
of
the
sites
in
our
network
and
the
accuracy
of
our
automated
filtering
is
supported
by
the
reactive
investigations
we
undertake
on
behalf
of
advertisers
who
contact
us
with
concerns.
We
have
found
through
reactive
manual
reviews
that
our
proactive
monitoring
misses
only
a
fraction
of
all
potential
invalid
click
activity
and
our
team
continually
works
to
improve
this
even
further.
5

7.
How
we
Help
Protect
Publishers:
The
relationship
between
Google,
advertisers,
and
publishers
is
based
on
trust.
As
a
result,
Google
treats
any
instance
of
invalid
activity
very
seriously.
In
order
to
help
protect
advertisers
and
publishers
from
invalid
click
activity,
Google
carefully
monitors
all
clicks
and
impressions
on
ads.
With
a
clean
network
we
preserve
the
advertisers’
trust,
which
means
better
earnings
for
publishers
with
clean
traffic.
We
understand
that
sometimes
a
third
party
may
generate
invalid
activity
on
your
ads
without
your
knowledge
or
permission.
We
know
that
these
clicks
are
not
the
result
of
genuine
interest
in
your
products
and
services,
and
regularly
work
to
improve
our
system
to
detect
these
clicks
and
filter
them.
Not
only
does
this
help
protect
publisher
accounts
from
sabotage,
it
also
ensures
that
advertisers
are
not
charged
for
the
filtered
clicks.
However,
AdSense
publishers
are
ultimately
responsible
for
all
of
the
traffic
to
their
ads.
As
a
publisher,
we
recommend
you
follow
our
policies
and
diligently
monitor
your
traffic
to
minimize
invalid
activity.
We
also
encourage
publishers
to
report
any
suspicious
activity
which
they’ve
detected
on
their
site.
6

8.
LITTLE
HISTORY,
HIGH-­‐PRICED
KEYWORD
CLICK
FRAUD
RISES
(APPENDIX
E):
The
Click
Fraud
Index
did
something
a
little
different
with
its
second-­‐quarter
report
on
click
fraud,
as
tracked
by
the
over
1,300
advertisers
and
agencies
participating
in
Click
Forensic’s
research.
They
split
out
the
click
fraud
rate
for
higher-­‐priced
search
terms,
priced
at
$2
and
higher.
Overall,
their
research
for
the
period
found
that
industry-­‐wide
click
fraud
had
increased.
The
first-­‐quarter
figure
of
13.7
percent
was
outpaced
by
the
second-­‐quarter’s
rate
of
14.1
percent.
Breaking
out
the
higher-­‐priced
keywords
showed
an
even
higher
increase.
Companies
participating
in
the
index
reported
click
fraud
at
a
20.2
percent
clip
for
those
keywords,
which
frequently
comprise
the
bulk
of
an
advertiser’s
spend.
“For
the
first
time,
we
have
industry
data
that
clearly
shows
what
many
have
expected
–
organizations
purchasing
higher-­‐priced
search
terms
are
significantly
more
vulnerable
to
click
fraud,”
said
Tom
Cuthbert,
president
and
CEO
of
Click
Forensics,
in
a
statement.
The
top
key
terms
averaged
$4.51
across
what
Click
Forensics
identified
as
the
top
five
search
advertising
industries:
Retail,
Financial
Services,
Health
&
Fitness,
Technology,
and
Entertainment.
That
was
down
from
$4.75
in
the
first
quarter
of
2006.
Rates
of
click
fraud
increase
dramatically
the
farther
one
gets
away
from
Google
and
Yahoo
for
search
advertising.
Click
Forensics
identifies
that
duo
as
Tier
1
search
providers.
For
the
second
quarter,
the
click
fraud
percentage
of
12.8
for
Tier
1
providers
was
a
little
higher
than
the
12.1
percent
reported
in
the
prior
quarter.
Tier
2
search
providers
(Ask,
MSN,
Lycos,
and
others)
had
a
click
fraud
rate
of
20.3
percent,
while
Tier
3
ones
(niche
sites
like
Dogpile)
racked
up
27.1
percent
for
the
second
quarter.
Both
figures
were
lower
than
their
first
quarter
percentages
of
21.3
and
29.8,
respectively.
Over
88
percent
of
click
fraud
takes
place
from
within
the
US
and
Canada,
the
report
noted.
Outside
the
US,
India
led
all
other
nations
with
unwanted
click
activity
rising
26
percent
during
the
quarter.
•
•
•
Disputes
over
the
issue
have
resulted
in
a
number
of lawsuits.
In
one
case,
Google
(acting
as
both
an
advertiser
and
advertising
network)
won
a
lawsuit
against
a
Texas
company
called
Auction
Experts
(acting
as
a
publisher),
which
Google
accused
of
paying
people
to
click
on
ads
that
appeared
on
Auction
Experts'
site,
costing
advertisers
$50,000.
Despite
networks'
efforts
to
stop
it,
publishers
are
suspicious
of
the
motives
of
the
advertising
networks,
because
the
advertising
network
receives
money
for
each
click,
even
if
it
is
fraudulent.
In
July
2005,
Yahoo
settled
a
class-­‐action
lawsuit
against
it
by
plaintiffs
alleging
it
did
not
do
enough
to
prevent
click
fraud.
Yahoo
paid
$4.5
million
in
legal
bills
for
the
plaintiffs
and
agreed
to
settle
advertiser
claims
dating
back
to
2004.
In
July
2006,
Google
settled
a
similar
suit
for
$90
million.
On
March
8,
2006,
Google
agreed
to
a
$90
million-­‐settlement
fund
in
the
class-­‐action
lawsuit
filed
by
Lane's
Gifts
&
Collectibles.
The
class-­‐action
lawsuit
was
filed
in
Miller
County,
Arkansas,
by
Dallas
attorneys
Steve
Malouf,
Joel
Fineberg,
and
Dean
Gresham.
The
expert
witness
for
the
Plaintiffs
in
the
case
was
Jessie
Stricchiola,
an
internet
search
expert
who
first
identified
instances
of
PPC
click
fraud
in
2001.
7

9.
REFERENCES
http://www.webpronews.com/highpriced-­‐keyword-­‐click-­‐fraud-­‐rises-­‐2006-­‐07
http://sponsoredlinx.com/blog/google-­‐adwords/preventing-­‐google-­‐adwords-­‐competitor-­‐click-­‐fraud/
http://en.wikipedia.org/wiki/Click_fraud
http://www.google.com/ads/adtrafficquality/invalid-­‐click-­‐protection.html
8