2006 Articles

Are You Ready for the "Next Big Thing" in Under 65 Individual Health Insurance Marketing?

This article appeared in the July 2006 issue of Healthcare Marketing Advisor

By Amy Houke

The landscape of today’s Under 65 health insurance marketplace is changing and that will have major implications for both heath insurance providers and health insurance marketers.  Employers, who have been — and continue to be — the primary source of health coverage for most Americans under the age of 65 are making concerted efforts to stem escalating health care expenditures. 

Consider the trends: A survey by America’s Health Insurance Plans (AHIP) of 99 member companies finds that as of March 2005, 71 offered Health Savings Account-eligible plans to large employers, 68 offered them to small employers, and 56 companies offer them to individuals.  AHIP also reports that the number of Americans covered by consumer-directed health plans (CDHPs) more than doubled in a six-month period to more than one million (March 2005).  Frost & Sullivan’s “2005 U.S. Consumer Directed Health Plan Markets Report” indicates that half the employers in the U.S. are considering a CDHP in order to cut down its health plan costs. 

The market is changing, the product mix is — of necessity — altering, and today’s targeting approaches are becoming outdated.  Health insurance marketers must adapt and change their approaches in order to stay competitive. 

Changing the Way You Market to the Under 65 Prospect Audience

Today’s Under 65 marketplace can easily be split into two general categories — the “Haves” and the “Have Nots.”  The “Haves” are: covered by an employer-sponsored or subsidized health insurance plan … retired and Medicare eligible … covered on a working spouse’s policy … or they are full-time college students who are covered on their parents’ policy.  The “Have Nots” typically fall into one of these scenarios: owner or employee of a small company … self-employed … part-time employees or students … unemployed … retired, but not yet Medicare eligible … recent college graduates … or recent high school graduates not bound for college. 

In the current environment, marketers have had good success reaching most of the “Have Not” segments through direct response marketing efforts — specifically direct mail.  Some “Have Not” segments are directly selectable — occupation questions on surveys serve to identify the self-employed and business owners — while some are inferred.  For example, we can assume that most people pursuing vocational training through distance education programs are likely in a part-time or low-level full-time position without benefits.  Most of the target segments are available via one of these selection methodologies on one or more list files commercially available for rental. 

The use of less-targeted, broader-reaching media can also be effective — free-standing newspaper inserts and direct response television, for example — by enabling the audience to self-identify.

Direct mail response rates in the Under 65 market average from 1.5 percent to 2.5 percent with broader print media pulling between .30 percent and .50 percent.  However, with a major change in audience behavior on the horizon, together with a new array of consumer directed product options, these traditional marketing approaches will gradually become less and less effective.

Audiences won’t be as comprehensively defined by their behavior or by their employment status, which requires that health insurance marketers establish an alternative means for targeting effectively.  Marketers of Under 65 individual health insurance products know that they have to stay ahead of the curve in their prospecting and sales efforts.  Most are wondering how. 

The utilization of credit data provides one key to addressing this shift.  The answer lies in modeling, more specifically, credit-based predictive modeling.  Many marketers can benefit from credit-based modeling as an effective way to predict consumer responsiveness and propensity to buy. 

Modeling — Defined

Modeling is a predictive and often multivariate technique, utilizing two cells of subjects: one containing individuals who have demonstrated a specific desired behavior (responded, bought, donated, converted, etc.) and one containing names from the entire pool that was solicited (e.g. the mail file).  A model aims specifically to compare the defining characteristics of those who responded versus those who did not, despite being given the same opportunity.  Algorithms are built that take into account several “key” variables and variable interactions, looking for those that are significantly represented among the group having demonstrated the desired behavior when compared to the other audience.  It then assigns “weights” or “scores” to each variable.  The heavier the weight, the more significant (meaningful) that variable is in predicting the desired behavior (response, conversion, etc.).  The output of a model is a scoring formula that can be used to segment a file into deciles, and to then select those segments of a file most likely to exhibit the desired behavior.

The Power of Credit Data

Credit data — credit scores as well as more than 300 individual credit attributes — are extremely indicative of consumers’ buying patterns, risk tolerance, lifestyles, and attitudes.  The use of credit data is far more effective than the use of demographics alone in predicting responsiveness and propensity to purchase.  Credit data was used to build a predictive response model for one of the nation’s premier marketers of individual Under 65 health insurance.  After a period of exclusivity ended, this model became available for use by other health care marketers.

This model, described in the following case study, was built using more than 14 million mailed pieces over the course of three years and has been updated with data from other health insurance marketers, including Blue Cross and Blue Shield plans.  The result is a model that transcends traditional targeting methodologies. 

Case Study

A major northeastern marketer of individual health insurance recognized the challenge they were facing in effectively reaching prospects for their Under 65 products.  They took advantage of the availability of a pre-built model and tested it for a new product they were launching — a high-deductible plan with an available Health Savings Account (HSA) component.  This product was very similar to the product for which the model was built and so model applicability seemed high.

The first test of the model was performed in a marketing region where the insurer enjoyed a strong and long-term presence.  Testing several list sources along with a selection from the credit-modeled list, the overall response rate generated was more than 2.00 percent, with the credit-modeled list pulling a strong 1.96 percent.  The credit data list was the second-highest converting list and a winner in terms of the response/ conversion blend used to assess a list’s overall profitability.  When the program was re-tested in a second marketing region where the brand was not as strong, the response was 1.75 percent.  Unfortunately it’s too early to assess conversion on this second effort.

So what did we learn?  At this point, traditional list selections (such as those mentioned above) designed to reach the “Have Nots” are still relatively successful.  However, one would anticipate the performance of modeled names to eventually surpass that of these vertical lists due to the changing audience profile.

The benefits of credit-based predictive modeling are many.  For example, there is a much greater potential rollout universe available, for a reasonable cost.  Response rates in the top deciles are sustainable compared to current, reachable response rates.  And this approach can serve as a single source of unduplicated names versus the current process of ordering multiple lists and losing between 20 to 40 percent in the merge/purge process.

While this model can serve as a “starting point,” if it performs well, marketers should consider building a custom model using the mail and responder files from their mailing efforts.  A custom model, built using demographics and infused with the powerful predictive capability of credit data, allows marketers to evolve with a changing marketplace. 

The future of targeting in a changing Under 65 environment requires the ability to adapt and to test new approaches such as credit data.  For some health care marketers, the opportunity represents a brave new world.  Are you ready?

About Amy Houke
Amy Houke is director of list and space media services at DMW Worldwide, a full-service direct response advertising agency with offices in Wayne, Pa, Plymouth, Mass, and St. Louis, Mo. The agency provides strategic planning, creative, database management, broadcast, media, production, fulfillment, and Web site promotion. This article is based on a presentation made at the 14th Annual Direct Marketing Conference for Independent Blue Cross and Blue Shield Plans. Houke can be reached at 314-432-3286 or via e-mail at ahouke@dmwdirect.com. Visit DMW on the Web at www.dmwdirect.com.