The document discusses Länsförsäkringar's journey from product-oriented to customer-oriented marketing by using scoring models and an offering engine. It describes how the offering engine calculates the probability and value of potential purchases for each customer to prioritize offerings. The scoring models assess multiple data points for each individual customer rather than customer groups to more efficiently target offerings. This approach has improved decision making and allowed Länsförsäkringar to develop customized, priority offerings for its entire customer base through its CRM system on a daily basis.
The journey from target groups to target individuals
1.
2. Which customers are we processing?
The journey from choosing target groups to selecting target individuals
Håkan Persson
Chef Affärsservice Privat
Länsförsäkringar
3. Description of the journey...
From product orientation to customer orientation
From basing the process on what we want customers to do, to basing it on what
customers is most likely to do
From inside out, to outside in
Make the organisation understand and utilise the benefit
4. Scored customers and an offering engine
First and foremost, which of our customers has the greatest potential to purchase
product X
Followed by, when and how shall we offer product X to the customer with the
greatest potential to purchase
5. Operative – our method?
The offering engine is based on the probability of purchase
The offering engine is presented in our CRM system, through the Yellow Tab
Communicate and work with the Yellow Tab at all customer meetings
Siebel
6. Scoring
A way to predict an event using statistical methods and historic information at
customer level
Used in many areas in society
Credit risks (Basel II), quality controls in the industry, etc.
Used in insurance
Claims risks, Fraud investigations, Portfolio clearing, Marketing, Premium
setting, etc.
The various methods assess the background information in relation to its
usefulness in finding the target data
E.g. The most important element for finding future purchasers of X is to have a Y
8. The Decathlon Principle
Although you may not be the individually best in your event, you still have a chance of winning
win since the results from several events are added together
In the same manner, there are several types of
background information about customers that
may contribute to the probability of purchase
By finding out the sub-totals of “each event”,
we will be able to secure more customers
You can view it as getting 3 points for a car,
4 points for a bank account and 2 points
for home insurance
If we set the qualifying limit at 4 points, car and
home insurance may be sufficient and we will
secure many good customers
If we set the qualifying limit on bank account,
we will not secure as many, although they all
have 4 points
10. What re we looking for...?
The goal here is not to make perfect decisions and find the definite buyer
The goal is to make a better choice than a random one, as often as possible
We can multiply our chances
But, the process may move from 0.5% to 3% in the group with which we are
working
And radically reduce the costs
One sixth
But the experience for the seller may not be a dramatically better choice
11. Result of scoring in a DR activity
Well-scored customers purchase more frequently
There is a proven effect of the campaign
Actual purchase frequencies since campaign start
Forecast and outcome per decile in the campaign group
4,50%
4,00%
3,50%
3,00%
2,50%
outcom
2,00% e
1,50%
1,00%
0,50%
0,00%
0,00% 2,00% 4,00% 6,00% 8,00% 10,00%
Forecast purchase frequencies 1 year
12. New sales of mortgages in practice
Process of offering engine
Option Probability Assessed Comparable Selected
Selection limits
Qualify the customer Calculate the probability of Calculate the anticipated Rank the business
for the offering purchase value potential
[e.g. 20% best]
13. In practice, choose the offering for the customer, not the
opposite
Mortgage
P=0.004
SEK=3000
Anticipated value SEK 12
Home owner Home owner
Home owner
P=0.05 P=0.05
P=0.05
SEK=300 SEK=300
SEK=300
Anticipated value SEK 15 Anticipated value SEK 15
Anticipated value SEK 15
Car
P=0.10
SEK=300
Customer already has the product
14. Calculations at customer level to support various business
objectives ..... what you want to achieve
•Probability of purchasing more of
existing products •Probability of purchase
Additional sales in Higher new sales of
existing commitment a certain product
Increased loyalty and
lower number of •Probability of cancellation
cancellations
15. We can work with all three types of offerings in the same
prioritisation...
Utökningserbjudande
Utökningserbjudande
• ”För att fullt ut utnyttja ditt skatteavdrag för
pensionssparande utnyttja öka skatteavdrag för
• ”För att fullt ut kan du ditt ditt månadssparande med 300
kr. Gör du det innankan du öka ditt månadssparande med 300
pensionssparande årsskiftet så bjuder vi på en julklapp”
kr. Gör du det innan årsskiftet så bjuder vi på en julklapp”
• Vi antar ett täckningsbidrag första året om 180 kr på den
affären ett täckningsbidrag första året om 180 kr på den
• Vi antar
affären
• Sannolikheten för köp beräknas till 180 kr * 8% = 14,40 kr
• Sannolikheten för köp beräknas till 180 kr * 8% = 14,40 kr
Nyförsäljningserbjudande Lojalitetserbjudande
Nyförsäljningserbjudande Lojalitetserbjudande
6
• ”Just nu har vi ett förmånligt erbjudande för dig som 6 • Under nästa år ger vi dig en självriskcheck på
försäkrar din bostad förmånligt erbjudande för dig som
• ”Just nu har vi ett genom Länsförsäkringar” personbilsförsäkringen värden självriskcheck på
• Under nästa år ger vi dig X kronor
försäkrar din bostad genom Länsförsäkringar” personbilsförsäkringen värd X kronor
• Sannolikheten att just den här kunden köper en • Sannolikheten för avhopp är 15%
bostadsförsäkring under den kommande året genom LF är
• Sannolikheten att just det här kunden köper en • Sannolikheten för avhopp är 15%
• Marginalen om kunden stannar kvar är ca 300 kr
bostadsförsäkring under det kommande året genom LF är
exempelvis 7% • Marginalen om kunden stannar kvar är ca 300 kr
exempelvis 7% • Affärsmöjligheten är värd 15% * 300 kr = 45 kr
• Täckningsbidraget för en boendeförsäkring antas vara ca 286 • Affärsmöjligheten är värd 15% * 300 kr = 45 kr
• Täckningsbidraget för en boendeförsäkring antas vara ca 286
kr på ett år
kr på ett år
• Här har vi en affärsmöjlighet värd 280 kr * 7% =20 kr
• Här har vi en affärsmöjlighet värd 280 kr * 7% =20 kr
4 7
4 7
16. Operative: work with the bonus concept and the offering
engine
The offering engine is based on the probability of purchase
The offering engine is presented in our CRM system, through the Yellow Tab
Communicate and work with the Yellow Tab at all customer meetings
Siebel
17. Organisational progress
A few years ago
Categorical customer selection (groups)
Long decision-making processes and arbitrary guesses
Inside out
Product orientation
Today
16 prioritised offerings updated daily
The selection includes all customers
We choose both customers (individuals) and products, simultaneously
Experience-based, assisted by scoring models
Customer orientation
18. Advantages of the scoring models
We select individuals, not groups of customers
Every customer has more than one chance of ending up in the target group
A more efficient way of using customer information
A reflection of the Länförsäkringar Alliance’s collective expertise for one year
We can simultaneously take into account the possibility of an event and the value
of the event
An excellent decision-making and prioritisation basis
Customer information can be assessed
Customer information can be used systematically and calculated monthly for each
individual
Can be used proactively and reactively
19. Conclusion
Scoring through predictive analyses
Offering engine
Provides an opportunity to develop the offering that is best-suited to our customers
every day
This is how we work, our entire customer base is assessed and qualified for our
prioritised offerings daily and presented through our customer system - Siebel
20. Future
Predictive models for irrational purchasing decisions
Connect external data to a larger extent
Internal training and further development – a continuous process