Customer
Profitability
Analytics
Faced
with
a
host
of
competitive
forces
and
their
accompanying
pricing
pressures
it’s
not
enough
just
to
grow
anymore,
a
financial
institution
must
grow
profitably.
The
key
to
growing
profitably
is
to
develop
business
tactics
and
strategies
around
customer
level
profitability
analytics,
including
the
lifetime
value
of
each
customer.
A
popular
definition
of
marketing
describes
such
business
tactics
and
strategies
as:
“The
science
and
art
of
finding,
retaining,
and
growing
profitable
customers.”1
For
this
article,
we
will
define
customer
profitability
as
the
sum
of
the
net
profit
or
loss
of
each
of
the
customers’
accounts.
Therefore
account
level
profitability
analysis
is
the
building
block
upon
which
to
analyze
customer
profitability.
It
is
also
the
building
block
to
analyze
product
profitability,
line
of
business
profitability,
branch
profitability,
regional
profitability,
and
ultimately
organization
profitability.
Part
1.
Customer
Acquisition
&
Origination
Costs
The
cost
to
acquire
a
customer
varies
by
context,
but
can
generally
be
calculated
as
the
cost
to
generate
a
response
divided
by
the
response
rate2.
For
example,
let’s
assume
you
are
going
to
try
and
acquire
new
customers
using
traditional
direct
mail,
such
as
a
postcard.
The
first
option
is
to
randomly
send
postcards
to
a
specific
area.
The
next
option
is
to
send
postcards
to
a
targeted
list.
Let’s
assume
that
the
random
option
costs
$.50
and
yields
a
1%
response
rate.
Let’s
assume
the
targeted
option
cost
$.70
and
yields
a
4%
response
rate.
Acquisition
cost
=
cost
of
sending
postcard/response
rate
Random
Option=.5/1.0%=$50.0
Targeted
Option=.7/4.0%=$17.5
In
this
example
it
is
clear
that
using
a
target
list
is
worthwhile,
despite
the
increased
initial
cost.
However,
the
key
point
is
not
whether
to
use
a
target
list
or
not,
the
key
point
is
to
measure
the
cost
of
acquiring
a
new
customer.
In
addition,
acquisition
costs
should
include
administrative
expenses
related
to
existing
customers
acquiring
new
products,
such
as
origination
costs,
which
could
be
calculated
as
a
set
fee
amount
or
a
percentage
of
original
balance.
Part
2.
Cost-‐to-‐Serve
Analysis
1
Philip
Kotler
and
Gary
Armstrong,
Principle
of
Marketing,
Prentice
Hall,
2001
2
Customer
Profitability
and
lifetime
Value,
Elie
Ofek,
Harvard
Business
School
Publishing,
copyright
2002
In
the
world
of
banking
the
cost-‐to-‐serve
variable
is,
well,
quite
variable.
Some
customers
may
put
extreme
pressure
on
certain
channels
such
as
branch
visits,
while
other
may
hardly
require
any
cost
to
serve
at
all.
Cost-‐to-‐serve
metrics
can
be
calculated
as
the
channel
multiplied
by
the
cost
of
using
the
channel3.
For
example,
#
of
branch
visits
x
cost
per
visit
#
of
calls
x
cost
per
call
#
of
atms
transactions
x
cost
per
transaction
Etc.
=Cost
to
Serve
In
addition,
certain
products
require
accounting
for
maintenance
expenses,
which
can
usually
be
expressed
as
a
percentage
of
outstanding
balance.
Part
3.
Customer
Break-‐Even
Analysis
Once
the
acquisition
and
the
cost-‐to-‐serve
costs
have
been
calculated,
the
next
step
is
to
perform
a
break-‐even
analysis
to
determine
the
sales
activity
required
to
realize
a
profit
from
the
customer.
Account
revenues
can
generally
be
calculated
as
net
interest
margin
x
balance
+
fees.
However,
we
also
need
to
include
a
risk
component,
which
can
be
expressed
as
a
provision
amount
and
a
charge
off
amount.
One
could
use
probabilities
multiplied
by
balances
or
just
prorated
amounts
for
provisions
and
charge
off
expenses.
Taking
into
account
revenues,
expenses,
and
risk
we
have
the
following
break-‐even
equation:
(Net
Interest
Margin
x
Balance)
-‐
Loan
Loss
Provision
-‐
Charge
off
=
Acquisition
&
Origination
Costs
+
Cost
to
Serve
One
can
then
readily
calculate
the
margin
required
to
breakeven
by
solving
for
required
Net
Interest
Margin
as
follows:
Net
Interest
Margin
=
(Acquisition
&
Origination
Costs
+
Cost
to
Serve
+
Loan
Loss
Provision
+
Charge-‐Off)/Balance
The
math
is
easy.
The
difficulty
lies
in
accurately
calculating
the
equation
components,
especially
cost
to
serve
expenses.
Part
3.
Lifetime
Value
Analysis
The
value
of
a
customer
is
more
than
just
the
initial
revenue.
A
lifetime
value
analysis
provides
a
prediction
of
the
net
profit
attributed
to
the
entire
future
relationship
with
a
customer.4
An
unprofitable
3
Evalution
of
Customer
Profitability
Analytics
Vendors,
Celent,
September
2007
customer
in
the
first
few
years
may
turn
out
to
be
a
profitable
customer
in
future
years
and
consequently
have
a
positive
lifetime
value.
Calculating
a
lifetime
value
for
each
customer
has
the
benefit
of
quantifying
the
long
term
relationship
with
the
customer
and
providing
insight
into
developing
optimal
strategies
for
each
customer.
For
example,
you
might
decide
to
“fire’
certain
unprofitable
customers,
reward
others,
and
identify
profitable
cross
sale
opportunities.5
Inputs
needed
to
calculate
a
lifetime
value
include:
retention
rate,
discount
rate,
contribution
margin,
retention
costs,
and
the
time
period.6
! ! ! (1 + ! ! !)!!.!
+ )!
= × (1 − ×
!!! !!!
CLV=Customer
Lifetime
Value
GC=Annualized
Gross
Contribution
M=Relevant
Retention
Costs
Per
Customer
Per
Year
r=
Annualized
Retention
Rate
d=Annualized
Discount
Rate
n=Time
Horizon
(in
years)
A
simplified
version
that
assumes
long
lasting
values
for
contribution
margin,
retention
rate,
and
discount
rates:
= ×(1 1+ )
+ −
Of
course
customer
profitability
analytics
is
easier
said
than
done;
easier
to
compute
in
an
academic
sense
than
in
reality.
However,
it
is
important
to
recognize
that
growth
and
market
share
does
not
equal
profitability,
so
a
financial
institution
needs
to
grow,
but
grow
profitably.
In
order
to
grow
profitably
it
must
measure
profitability.
The
best
way
to
measure
profitability
is
via
the
lifetime
value
of
the
customer,
which
relies
on
accurate
account
level
profitability
accounting.
3
www.wikipedia.com,
Customer
Lifetime
Value,
7/9/12
5
Evalution
of
Customer
Profitability
Analytics
Vendors,
Celent,
September
2007
6
www.wikipedia.com,
Customer
Lifetime
Value,
7/9/12