Increasing Insurance Customer Retention
by Arthur Middleton Hughes


Most insurance companies are set up for new customer acquisition.  Acquisition is easy to measure.  Bonuses are paid for new customers. You can count the new customers on your database. Customer retention is harder to measure. How can you prove that, as a result of your efforts, 42,687 customers did not drop their coverage last year? This is tough to prove, and therefore hard to reward.

At the same time, retention can be very cost effective. Many studies have shown that $1 paid towards customer retention increases profits by more than $5 spent on new customer acquisition. What are companies doing about this situation?

This article covers two programs which not only reduced customer attrition, but were able to prove it to managementís satisfaction. The two methods were:

  • Sell a second policy to existing customers.
  • Communicate often with customers.

 Selling a second policy

Most insurance companies try to sell an additional policy to their current policyholders. What they may not realize is that the value of the second policy is much more than the additional profit from that policy. The value is in increasing the retention rate (the persistence) of the first policy. Somehow, when people own two (or more) policies from a company, they are more loyal than if they own just one. Banks have discovered this, and many of them have been able to create graphs like this:


Retention rate based on the number of products owned.


Using this knowledge, several companies have launched organized programs to get their agents to sell a second policy by offering customers a discount on their current policy if they buy another product. The most advanced programs use the web as a delivery vehicle to the agents.

First, the company uses their database to determine the Next Best Product (NBP) for each customer, based on what they know about the customer, and the typical buying patterns of customers with similar demographics. Next, they figure out the lifetime value (LTV) of each customer now and later if they were to buy the next best product. Letís assume that the present LTV after five years is $530 and, if they were to buy their NBP would be $930.  This means that the company will gain $400 in profit by selling the additional product. From this you can subtract the cost of selling the new product, letís say $100.  If you were to offer the customer a $100 discount on the current product as an inducement, then you would still be $200 ahead in profits.


 5 Year

Current LTV $530
LTV with NBP $930
Gain $400
Marketing Cost $100
Premium $100
Net Gain $200

The Next Best Product software scores every customer by propensity to purchase, ranking them in deciles from most likely to least likely. These scores can be used to determine which customers to mail to on a rollout.

How to go about selling the second product.

Armed with your customer database with Next Best Product calculated and placed in each customer record (along with the saving that you can provide as a premium for the NBP purchase) you set up a web micro-site so that all of your agents can see their existing customers and their next best products.

Each agent logs on to your web micro-site and views a list of customers, their NBP and the dollar discount available to them on their current policy. There is a draft direct mail letter to each customer which the agent can view and edit on the web. He can change the customerís name to a nickname, for example.  He can eliminate customers from the list who he knows will not buy the second product. Then he clicks send.


Really good letter text includes a yellow lift note from the agent to the customer saying something like: ďDear Bob and Kitty: According to my calculations, it looks like your household could qualify for savings of about $100 on your auto premium. Call me today!  Susan.Ē

For those who do not respond, a second letter can go out a month later with a reminder notice. In addition, the web micro-site provides the agent with the phone numbers so he can follow up then and there without having to look the numbers up.  The web site should make it easy for the agent to report feedback by clicking boxes next to each name such as:

  • Appointment

  • Not Interested

  • Visited with the customer

  • Made phone call

  • Policyholder purchased new policy.

To make this work, of course, the biggest difficulty to overcome is to sell the system to the agents.  Each agent has their own methods. Some are adverse to using a computer. An organized campaign to sell the program has to be developed and carried out.

The other requirement is the need to set up a control group of customers who do not get the letters. Some of them will buy the Next Best Product without any letter, because of various other promotions that are always being conducted. What you have to prove is whether these letters (and the discounts offered) were working. Only by comparing the results with a control group can you prove to management that your money is well spent.

One company that conducted a program designed along these lines managed to get more than 80% of the agents in a test area to buy in to the program. The letters worked. Sales of the Next Best Product to those who received the letters were more than ten percent higher than sales to people who did not get the letters.  The predictive model worked extremely well. The top decile had a sales rate of almost 70% higher than the average. The follow up phone calls increased the sales rate by more than 40% higher than those who did not get a follow up phone call. Finally, the retention rate of those who purchased a second policy due to the program was more than 6% higher than the rate of those who did not purchase a second policy.

Communicating with customers

Home and automobile insurance is a tough business. Because of vigorous competition and high acquisition costs, it takes several years before an automobile insurance customer can become profitable. If the customer leaves a year or two after being acquired, the insurance company loses money. Travelers knew this but was not able to do much about it. Travelers works through thousands of independent agents who handle insurance for many different companies. A database analyst from CDC set up a communications system for Travelers that worked. From previous experience the analyst knew that:

  • Similar programs attempted at Travelers had not been successful, however,

  • Customers want communications. They like to hear from their insurance agents.

  • To be effective, the communication should come from a local agent, not from a national headquarters.

  • A 1% increase in the customer retention rate would be worth millions of dollars in increased annual profits to Travelers.

  • To get this type of increase in retention, the program would have to get 15% of the agents covering 25% of the customers to buy into the plan.

To begin, it was necessary to sell the program to the independent agents. It met a lot of resistance. They were not convinced:

  • That a retention program was valuable to them

  • That direct mail would work

  • That communications could affect when or whether a customer will defect

  • That the program would have an adequate ROI

  • That Travelers knew anything about their customers that they didnít already know

What was finally developed was a retention program, built from a customer database that provided a systematic program delivering high quality communications at a very low cost. The messages were from the local agent. The program provided the agents with a turnkey operation which was simple to buy into, and required almost no work on the part of the agents themselves.

The program developed five annual "touches" which varied with the type of insurance that the customer had, and the length of time that the customer had been with Travelers. The five were:

Within 60 days of renewal

An annual review of the policy

Within the 1st quarter

A thank you card

Within the 2nd quarter

A cross-sell postcard

Within the 3rd quarter

A newsletter

Within the 4th quarter

A seasonal greetings card

The program developers learned that for each customer, they had constantly to determine the appropriate message, the frequency of messages that the customer wanted, the desired channel, the timing of the message, and the likelihood of defection. Statistics that showed that 65% of the customers who defected, never talked to an agent before they left. But 80% of the customers who talked to an agent during the year did not leave.

What do customers want from their independent agents? A Customer Retention Survey for Personal Lines Customers conducted by the Independent Insurance Agency Association showed that:

  • 52% of insurance customers describe themselves a relationship buyers

  • The customers want an annual review of the coverage of their policy

  • They are looking for an agent with integrity who has a stable business

  • They want information about their policies and coverage

The program was based on detailed analysis of the customer database. The program used database data and modeling to determine who was staying and who was leaving. The analysts determined customer profitability and lifetime value. These were used to drive the segmentation and retention strategy. Overall, the program came up with a measurement of customer desirability. Now that the software showed who Travelers wanted to keep, and what they were worth, it was possible to develop and execute a program to modify customer behavior through communications.

To get the agents to sign up for the program, it was necessary to make it easy for them. They were sent them a kit describing it. The got an 800 number to call, and were recommended a standard package. They had a Website for them to review the status of their program, and provided regular reports. To sell agents on the program today, the program showed them what happened to agents who bought the program in past years, vs those who did not. Agents who did not participate in the program lost 17.3% of their customers in the first year. Participating agents lost only 12.2%. Numbers like this proved that there was a good return on the agentís marketing dollar.

So what did the program accomplish? To measure the success, Travelers compared the retention rates of participants and non-participants. For auto insurance customers, the program was able to increase the retention rate by 5%. For property insurance, the increase was 4%. The charts look like this:

Why did this program succeed whereas the others had failed?

  • The program was not centrally subsidized

  • In prior programs, the agent risked nothing

  • In this programs, the agent was risking his own money. He wanted it to succeed.



Arthur Middleton Hughes is Vice President of The Database Marketing Institute. Ltd. ( which provides strategic advice on relationship marketing. Arthur is also Senior Strategist at ( which provides precision e-mail marketing services for major corporations worldwide. Arthur is the author of Strategic Database Marketing 3rd ed. (McGraw Hill 2006). You may reach Arthur at (954) 767-4558 .

The articles on this web site are available to the general public to read, enjoy and for limited business use. If you want to reprint more than one or two of them for resale or use in a business or educational environment, send an email to Arthur Hughes at He will give you permission by return email. The cost, depending on the number of copies you want to reprint, is very inexpensive.