32. >
KPIs
Depend
On
Solid
Founda=on
Business
Objec.ves
1
KPIs Web
2
Analy.c
3 Data
Founda.on
Key
Performance
Indicators
also
depend
on
a
solid
founda.on
of
well-‐
defined
page
names,
content
hierarchy,
and
report
suite
architecture.
Without
these
building
blocks
in
place,
making
decisions
and
taking
ac.on
on
the
data
will
prove
difficult.
33. >
User-‐friendly
Page
Names
§ Defining
a
good
page
naming
strategy
is
one
of
the
most
important
steps
in
Three
C’s
of
Effec.ve
Page
Naming
maximizing
Web
analy.cs
success.
Context:
Include
directory
structure
or
content
hierarchy
in
the
page
name
to
help
users
orient
§ In
order
to
help
people
understand
the
the
page
within
the
site
and
simplify
report
filtering.
performance
of
site
content,
TA
should
consider
crea.ng
more
user-‐friendly
Clarity:
Ensure
the
page
name
is
clear
and
easily
iden.fiable
for
infrequent
users.
page
names.
Conciseness:
Keep
the
page
name
as
short
as
§ You
will
need
to
create
page
names
possible
to
maximize
limited
character
space.
that
are
contextual,
clear,
and
concise.
34. >
Structure
of
a
Quality
Page
Name
Will
users
know
what
page
this
is?
Clarity
is
an
overarching
concern
for
the
en.re
page
name.
Clarity
Which
Context
neighborhood?
A
Friendly
Page
n
the
URL
Context
focuses
o Name
has
two
parts:
URL
structure
structure
stem,
which
helps
directory:subdirectory:sub-‐subdirectory:specific
page
name
stem
and
where
a
page
to
iden.fy
specific
page
resides.
name.
URL
structure
stem
Conciseness
Use
“US”
instead
of
“United
States”?
Conciseness
primarily
focuses
on
making
the
URL
structure
stem
as
short
as
possible.
The
specific
page
name
part
should
also
be
as
concise
as
possible
but
most
of
the
emphasis
will
be
on
the
stem.
36. >
Content
Hierarchy
Site
Sec.on
Level
Page
Type
Different
aggrega.ons
of
content
data
will
allow
you
to
iden.fy
Sub
Sec.on
Level
key
paOerns
at
higher
levels
and
then
drill
into
specific
details
at
lower
levels
Page
Level
37. >
Page
Naming
Examples
Good
Page
Naming
vs.
Bad
Page
Naming
§ Clear and user-friendly § Unclear and confusing
§ More concise § Too long (URLs)
§ Easy to filter and search § Awkward to filter and search
§ Consistent format § Inconsistent format
39. >
Page
Naming
Op=ons
Recommended
1. Server-‐side:
Use
server-‐side
logic
to
populate
page
name
for
each
web
page.
2. Hard-‐code:
Manually
set
page
name
on
each
web
page.
Use
With
Cau2on
3. PageName
plug-‐in:
JavaScript
plug-‐in
strips
“hOp://
www.domain.com/”
from
URL
page
names.
Not
Recommended
4. Leave
blank:
SiteCatalyst
defaults
to
page
URL.
5. Document.=tle:
Uses
the
.tle
of
each
page
instead
of
URL.
41. >
External
Campaign
Tracking
§ Campaigns
demand
very
special
aOen.on
within
a
global
type
structure.
– Dont
duplicate
tracking
codes
for
disparate
campaigns
– Use
structured
tracking
codes:
cid=e:tar:0003
§ As
a
best
prac.ce,
we
recommend
crea.ng
uniform
tracking
codes.
Examples:
All
Marke.ng
– cid=a:033007
à
“a” flags
affiliate
campaign
– cid=e:033007
à
“e” flags
email
campaign
– cid=sem:g:033007
à”sem:g” signifies
Google
and
is
a
paid
search
campaign
Affiliates
Other
§ U.lize
SAINT
to
upload
valuable
meta
Emails
Redirects
data
for
analysis
Online
Marke.ng
45. Product
Classifica=ons
§ Product
pages
are
a
key
focus
of
repor.ng
and
analysis.
12345
§ Populate
the
s.products
variable
with
all
product
IDs
viewed
on
a
page.
§ In
order
to
gain
more
insights
into
these
important
pages,
you
can
leverage
SiteCatalyst’s
SAINT
classifica.on
tool
to
upload
metadata
with
67891
different
product
aOributes
into
SiteCatalyst.
§ Able
to
analyze
the
conversion
performance
of
its
product
pages
by
aggregated
product
aOributes
such
as
category,
subcategory,
region,
facili.es,
23456
accommoda.on
type,
star
ra.ng,
etc.
§ All
meta
data
is
.ed
to
the
specific
product
id
s.products=";
12345,;
67891,;
23456"
52. >
SAINT:
Campaigns
Classifica=ons
XID
§ By
u.lizing
the
SAINT
classifica.on
template,
you
can
AOributes
con.nue
to
upload
meta
data
for
each
variable
§ Meta
data
examples:
SAINT
– Campaign:
Name,
Channel,
Owner,
Paid
vs
Nonpaid,
Branded
keywords
vs
non-‐branded,
etc
– Product:
Name,
Category,
Brand,
etc
– Customer
ID:
Profitability,
segment,
churn
risk,
demographics,
loca=on,
etc
§ The
process
of
assigning
aOributes
through
SAINT
can
be
automated
via
FTP
§ Classifica.ons
can
be
updated
at
any
.me
and
will
change
all
data
retrospec.vely,
because
they
are
a
label
and
do
not
change
the
underlying
variable
value
and
the
data
recorded
against
it.
69. PageNaming
Op=ons
–
Pros
/
Cons
Pros
Cons
Server-side — Provides flexibility to create — Only applies to dynamic pages
friendly page names — Flexibility depends on platform
— Logic automates page — Ongoing diligence to ensure logic doesn’t break in future
naming process
Hard-code — Free to create friendly page — Can be labor-intensive
name that isn’t bound by — Doesn’t apply to dynamic pages
URL or page title
— Must be careful when using pages as boilerplates
Page Name — Requires less effort to create — Only recommended for sites with well-defined URL
page names structures
Plug-in — Shortens URL page names — Can still lead to long page names
and makes them more — Won’t aggregate a page with different URL variations
manageable
— Problematic with dynamic pages
Leave Blank — Requires no effort — URL page names can be long and unwieldy
— Limited character space is wasted on domain root
(defaults to URL)
— Won’t aggregate a page with different URL variations
— Problematic with dynamic pages
Document.title — Requires little effort to set — Page titles aren’t always unique to a page
up — Can be long and contain unnecessary keywords (SEO)
— May change frequently (SEO)
— Problems caused by translation tools (e.g., Babel Fish)