WHEN THE CUSTOMER WRITES HIS

OWN STORY
A SEGMENTATION SCHEME FOR THE
LIFE INSURANCE MARKET

Jomar F. Rabajante and Allen L.

Nazareno
University of the Philippines Los
Baños
Life insurance is not just about
business, it is also personal.

Life insurance products are not

“bought” but are “sold”.
The Segmentation Scheme
Agen
t

Marketing
Departmen
t
helpful in customer profiling in two ways:
Product Matching and Semi-Personalized
Customer Care.

useful in enhancing renewal payments (life

insurance does not lie on the 1st -year premium
payments alone)

useful in the follow-up marketing activities such

as up-selling and cross-selling of other
insurance products
let the clients write their own story with
agent’s guidance

make a concept map by starting with a

central concept and then branching out
to more, related concepts
quantify qualitative decision making

capture the uncertainty in human

judgment

used to decide which priorities (called

as criteria) should be ranked first,
second, third, and so on
TABLE 1. Saaty’s Scale for Pairwise
Comparison
The comparison matrix A, which follows
on the client’s judgment between the
relative importance between two
factors, is constructed.
The matrix above is transformed into a
fuzzy comparison matrix with a
different membership scales
TABLE 2. Scale for Fuzzy Pairwise
Comparison
The scales used in the table follow the
relationship:
aij
rij =
aij + 1
The priority weights are determined
using the following formulas:
bi
wi =
 n 
 ∑ bi 
 j =1 
where,
1
bi =
 n 1
∑  − n
 j =1 rij 
 
The consistency of the generated
comparison matrix is determined using
the consistency ratio:
CR = (Consistency Index of matrix A)
(Random Consistency of matrix A).
  0.283905 

W = 0.336879  
0.379216 
new technique in grouping the customers

demographic and psychographic profiles are to

be considered, as well as the relatedness of the
stories of the customers

integration of all the outputs from the mind

map and Fuzzy AHP

other than customer profiling, the insurance

company can also determine what kind of
customers the agents tap
Robust Product Matching

Characterist
Prod ic W Prod
uct D uct B
Characteristi
Characterist cX
Character ic V Characteris
istic U tic Z
(Unique
for
Product
D) Characterist
Prod ic Y Prod
uct C uct A
Semi-Personalized Customer Care
Customer Customer Customer
1 2 4

OFFER
MARKETING
OFFER A connection ACTIVITY II
MARKETING means that the
ACTIVITY I two customers Customer
have the same, 5 Customer
let say, priority 7
J.

OFFER MARKETING
ACTIVITY I, II or III or
Customer 1/Rank of Char J = 1/2 Customer none
3 6

(1/Rank of Char K = 1/ 4) + (1/Rank of Char L = 1/3)

Outside underwriting considerations, the
customers can be grouped based on the
combination of factors, which includes:
demographics (age, gender, civil status,
income, education, profession/occupation,
location, family size, religion, nationality, etc.)
and psychographics/lifestyle (spending
patterns, future plans, brand preference,
priorities, etc.).

The choice of insurer based on the price of

insurance policies is not a major concern for
customers since the prices are usually uniform
in the industry (Philippine setting).
Artificial Neural Networks, Cluster Analysis and
other multivariate statistical analysis are
used for grouping customers. However,
Network Analysis is a new data mining
technique, which does not only give list of
groups but can also give visualization of and
flexibility in the grouping.

Customer profiles need to be updated

frequently, e.g. 2 to 3 years. Softwares
should be used for a large database.
The customer “stories” must be
heard, and this will help the
insurance industry spread the
benefits of having financial
protection.
 THANK
YOU.