next best product to sell gap
TRANSCRIPT
![Page 1: Next best product to sell gap](https://reader038.vdocuments.net/reader038/viewer/2022100723/58f155821a28ab58598b456f/html5/thumbnails/1.jpg)
1
Next best Product to Sell
Increase the “Per Customer Range” of Products in an Insurance Arena
New Developments in Measurement and Analytics
![Page 2: Next best product to sell gap](https://reader038.vdocuments.net/reader038/viewer/2022100723/58f155821a28ab58598b456f/html5/thumbnails/2.jpg)
2
Background
• Client : Major Life Insurance Company in Asia
• 12 years in Operations
• Product Range Included:
– Term Policies
– Unit Linked
– Endowment
– Pensions
• All polices had the option of Riders
![Page 3: Next best product to sell gap](https://reader038.vdocuments.net/reader038/viewer/2022100723/58f155821a28ab58598b456f/html5/thumbnails/3.jpg)
3
Task at Hand
• After 12 years of operations there were 17 million
customer on the books
• Over 95% of customers had only one policy
• Identify Customers who had the propensity to buy an
additional policy and /or Rider
![Page 4: Next best product to sell gap](https://reader038.vdocuments.net/reader038/viewer/2022100723/58f155821a28ab58598b456f/html5/thumbnails/4.jpg)
4
Technique
• Logistic Regression was used to discriminate between
those customers who had bought two policies and those
who bought just one
– Gender, Age Group, Nominee Relationship, Sum Assured,
Time since Last Purchase were variables that discriminated
between the two groups
• Every Customer had 4 probabilities computed for each of
the other products and for one additional rider
![Page 5: Next best product to sell gap](https://reader038.vdocuments.net/reader038/viewer/2022100723/58f155821a28ab58598b456f/html5/thumbnails/5.jpg)
5
Technique
Pictorial Representation
Customer Id:
1001943 TERM
POLICIES
UNIT LINKED ENDOWMENT PENSIONS RIDERS
0.9 0.7 0.8 0.4
Target this product based
on high propensity to buy
Next best
Product
Propensity
of Buying
Customer
Already Owns
![Page 6: Next best product to sell gap](https://reader038.vdocuments.net/reader038/viewer/2022100723/58f155821a28ab58598b456f/html5/thumbnails/6.jpg)
6
Implementation
• A Cross Sell Campaign was created targeting every
customer with the product which had the highest
probability of being bought
• Consisted of an Email with communication that spoke of
the benefit of the targeted product
• Two audiences were tested with the campaign
– those who were identified by the model for a specified
product
– A random selection of customers across all probabilities
![Page 7: Next best product to sell gap](https://reader038.vdocuments.net/reader038/viewer/2022100723/58f155821a28ab58598b456f/html5/thumbnails/7.jpg)
7
Results
• The test group (those with high probabilities) out pulled
the control group (the randomly selected) by a ratio of
1.9 : 1.0 on response rate
• They also bought higher value polices with an average
difference of $ 53 for the same type of policy
• The tenure of policies was not significantly different
between the two groups
![Page 8: Next best product to sell gap](https://reader038.vdocuments.net/reader038/viewer/2022100723/58f155821a28ab58598b456f/html5/thumbnails/8.jpg)
8
Bangalore, IN Office:No. 141, 2nd Cross, 2nd Main,Domlur, 2nd Stage, Bangalore 560071Phone: +91 80 40917572, +91 80 [email protected]
Contact Us US Office:
Suite 100, 1780 Chadds Lake Dr, NE
Marietta, Georgia, 30068-1608
Atlanta, USA