Imagine leaving your sales and marketing efforts to guesswork. Not understanding your customer's age, lifestyle, income or other factors would spell certain disaster for any insurance carrier trying to market or cross-sell specific products.Amica Life Insurance, a wholly owned subsidiary of Amica Mutual Insurance Company, is one company that decided to attack its marketing efforts with statistical intelligence.
Based in Lincoln, R.I., Amica started insuring policyholders in the northeast in 1907. Today, with 41 branch offices and more than 3,100 employees, Amica serves more than 639,000 households from coast to coast. As one of the oldest mutual insurers of automobiles in the U.S. and a national writer of automobile, homeowner, marine and personal umbrella insurance, Amica markets its life insurance products primarily to its auto and homeowners policyholders. Its primary cross-selling vehicle is direct mail or telephone campaigns.
Like many carriers, Amica Life Insurance has faced increasing competition. The entry of banks and other insurance companies competing for the same customers, along with the open door of the Internet to consumers comparing rates, lead the company to look at ways to increase its marketing effectiveness.
In the face of this competitive climate, marketing manager Ed Naya needed to find a way to sustain the increase the company had experienced in new insurance inquiries, and to do that he needed to develop a strategy for being able to better target its market, while at the same time sustain the increasing inquiries-without increasing the marketing budget.
Before approaching executive management about the plan, Naya gathered his marketing team to discuss the short- and long-term objectives.
"We knew predictive modeling would be key to meeting our marketing goals," said Naya. The company hired a consultant to build a single custom model, the "Inquiry/Purchase Model," to predict the propensity of a prospect to inquire and buy life insurance. The consultant worked with Naya and his team to create the idea of "deciles"-or customer segments-as a way to explain the concept of modeling to management.
The Inquiry Model identified 12 predictive variables that would most likely predict an inquiry and then created 10 deciles that the team tested and scored by likeliness to purchase life insurance.
After presenting his idea to senior management, Naya got the go-ahead to proceed. Implementation of the Inquiry Model required the hiring of a statistician, who tested the model with a random sample direct mail campaign and found gold: The 10 deciles responded as predicted.
That discovery lead Naya and his team to think even further outside the box: It required not one model, but a modeling application that could be used throughout the organization for numerous products, campaigns and data sources to create the most effective campaigns across the enterprise.
After evaluating various modeling applications, Amica Life decided on the Affinium Model from Unica, Waltham, Mass.
"The application is built specifically for marketing applications," Naya points out, "and can access numerous data sources without programming." Although the software offers sophisticated and flexible predictive models, it's easy to use, Naya adds.
The top five
Unica installed Affinium Model in two months, which included one week for actual installation and the remaining time for marketing staff training. Naya's team gained experience with the Affinium Model for three months by running level term insurance campaigns. Then the team used Model to score the entire customer base to create a "top five" segmentation ranking and expanded its modeling to additional products, including annuities, whole life and long term care campaigns.
In it's "Top Five" model, Segment A was the most likely to inquire and Segment D the least.
Attributes of "A" policyholders include: newly insured with Amica; have children; married; professionals; have had a positive experience with a recent claim.
Attributes of "D" policyholders include: single; no children; under age 25; not a homeowner; no recent claims.
The predictive modeling approach has allowed Amica to change who it targets and what it offers.
In the past, marketing dollars were evenly distributed across the entire target audience and resources were allocated equally among the elements of a direct mail campaign (package design, product or service offer and list segmentation).
Amica Life had also used single attributes-such as married or single-to select prospects for its mailings. Although being married remains a key variable in identifying who is likely to purchase life insurance, it is only one of several important criteria.
Since Amica Life has been targeting its best prospects, it's shifted its allocation of marketing resources and dollars. In the past, equal marketing dollars-20% each-would go to all five deciles.
Now those figures have changed: Amica spends more on segments A and B and less on C and D segments. For example, with the assistance of modeling, in the first year Amica mailed to Segment A five times during the year. This year they will mail to Segment A eight times. The result is a reduction of 25% in marketing costs.
Ultimately, the Affinium Model's ability to identify those customers most likely to inquire about and purchase Amica Life's products has helped Amica exceeded its annual goal of inquiries by 110%.
"We really want to sustain this level of activity," Naya says. "Previously we had a .5 to .75 reply ratio, and we've seen it jump to 1.2 to 1.5, holding steady."
Naya admits that the industry in general has seen an upturn. "Since Sept. 11, 2001, there has been an obvious increase in the number of customers considering life insurance," he says.
"But this system allows us to identify other trends under the surface that contribute to an individual's inclination to purchase, such as a recent claim that would signify a life-changing event."
The hallmark of successful direct marketing is having a great list, since it determines whether or not your offer is put in front of your best prospects.
"I've been in marketing for almost 20 years, and this reinforces what I learned back in graduate school," says Naya. "The right data is a powerful tool."
The carrier has kept on eye on improving its processes, i.e., it's incorporated the Affinium Model into its direct mail campaigns and reapplied an improved model each time campaign results come back. It also purchases additional demographic data from Claritas, San Diego, Calif., and appends it to its own system. As results are tallied, a likely responder's attributes are included in the next scoring run to update and fine-tune the segments.
Amica will continue to focus its modeling efforts on both the targets of its messages and the timing of its direct mail campaigns. Enhanced analysis of its customer base will help Amica further tailor the messages individuals within segments receive and when they receive them.
For example, some customers require life insurance to protect a spouse, but some simply want to ensure a comfortable retirement. Still others purchase life insurance as a long-term investment to help secure their children's education. These customers may be in the same segment, but they have different needs for life insurance, therefore, require different targeted messages.
Likewise, customers vary in the time chosen to buy life insurance. For some it's shortly after being married, for others before they retire or after they have their first child. Amica's goal is to remove the guesswork involved in timing its marketing efforts to coincide with these critical times, solidify customer relationships and keep competitors at bay in the process.
Register or login for access to this item and much more
All Digital Insurance content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access