Published: 06.21.2021



Measurement and Reporting

Real World Stories

How to Boost Regional Medicare Plan Results With Rigorous Campaign Testing Optimization

In our Anatomy of a Solution series, we take a look at some direct challenges and disruptive solutions that Amsive has driven for clients. This week, we look at challenging AEP assumptions by rigorous testing with surgical precision to drive acquisition growth.

The Basics

A large regional Medicare plan had used a set rollout campaign that included five direct mail touchpoints for years, but due to budget cuts, ultimately sought a way to reduce its marketing expenses while simultaneously driving aggressive Medicare AdvantageiMedicare Advantage is part of the Medicare program, but is offered to anyone age 65 and older by private insurance companies instead of the federal government. Also referred to as Medicare Part C, these plans typically include hospital, medical and prescription drug coverage. Medicare Advantage is different from standard Medicare, which is still available even with a Part ... Read More member growth. The ability to test new rollout options during the relatively short AEP window during this campaign led us to try to question industry assumptions on how channel and messaging strategies were done. A new rollout could be optimized for campaign efficiency — and we set out to prove it.

The hope was to be able to zero in on delivering effectively connected campaign experiences to those audiences across all channels. But finding the right strategies would be key to reducing cost per leadiCost Per Lead is a common type of performance-based advertising where an advertiser pays a digital publisher every time a lead is generated. This type of advertising model puts the responsibility in the hands of both the advertiser, who has to maximize the leads generated, and the publisher, who will be compensated based on the number of leads their ad generates. So it's i... Read More and per acquisition.

The Challenge

The Medicare client was in a bind — they had a set number of touchpoints across direct mail and digital display, but wanted to effectively reduce their budget to achieve their desired acquisition outcome for the AEP season. They wanted to figure out how much to whittle the cost on a set number of touchpoints, but the question was: which ones could achieve this budgeting and creative balance?

We had to build and test rollouts using a control group and shifting different mail drop and messaging cadence patterns.

The tricky first step on such a campaign is that effective Medicare services are broken down by county, so we set out to create modeled audiences on a county level. Filtering to that relatively granular level may seem easy, but remember — no two neighborhoods are the same. Take the New York City metro area, for example. Audiences in the city in the Bronx may not respond to the same type of strategy to audiences in Westchester outside of the city; audiences in the heavily populated city borough of Brooklyn may not respond to the same type of strategy as audiences on suburban Long Island. 

With that critical step in mind, we had to build and test rollouts using a control group and shifting different mail drop and messaging cadence patterns. The primary reason we suggested a testing strategy? Many clients had simply never done it before. They assumed a strict cadence worked and didn’t budge from that within the relatively limited AEP sign-up window. 

We set out to challenge that assumption and were confident in moving ahead bearing that out with our testing strategy.

The Solution

Developing a data-based audience strategy to create prospect modelsiA predictive model in marketing is a tool used to predict buyer behavior. A predictive model identifies customers or prospects who have the right demographic or psychographic behavior to rate them as someone who has a high propensity to purchase a certain product. The predictions can also be a great tool to aid in communication with certain prospective customers. at the county level meant using our in-house statistical tools to collate potential audiences. Such lookalikeiA lookalike model is a group of people who, based on digital data, have similar interests and habits as a company or brand's target audience. Lookalike modeling helps identify new prospects that behave a lot like current customers or prospects. By identifying and marketing to this group, engagement and conversion rates can be enhanced. audiences allowed us to test the channel and messaging strategy using a control group and two different messaging patterns to find the perfect mix of direct mail and digital advertising. The desired outcome was achieving the highest return on marketing investment.

The testing phase primarily looked at whether we could reduce the client’s preferred five direct mail touchpoints — which consisted of an initial pre-heat message before the annual enrollment period starting, then three drops of enrollment-specific messaging, and a last chance message before the enrollment window closed. 

The control group from the lookalike audience — who were randomly selected and sectioned off based on deciles — helped challenge that typical cadence. By taking away the initial message in one group, or limiting the messaging rollout to the first four drops only, we were able to crunch the numbers to figure out the optimized number of touches and the optimized timing of when to achieve the highest ROI.

The Results

With our understanding and definition of the key business outcomes in terms of three key metrics of cost per lead, average net lift from timing tests, and cost per acquisitioniAcquisition cost is the amount of money a company or brand spends to attain a new customer. Determining your customer acquisition cost can be done by dividing the cost of a marketing campaign by the actual number of customers you aquired. Knowing this number can help brands determine proper marketing spending habits and ultimately increase ROI., we determined one less mail touch with focused and efficient messaging achieved the highest return on marketing investment.

By using the testing cycle, we were able to increase results and yet still lower costs to give the Medicare plan provider:

  • A 16 percent reduction in cost per lead
  • An 88 percent average net lift from the timing test
  • A 5 percent reduction in cost per acquisition

For a mature Medicare client in an industry that rarely changes, this was huge. This data-centric approach, with key insights to determine which is the optimal time and cadence, challenged industry assumptions about channel and messaging strategy and created a repeatable blueprint for future cost savings and campaign alterations.