Analytics White Paper

Optimizing Messaging & Positioning With Choice Modeling

Messaging and positioning choice modeling is recommended when the primary research objective is to obtain information that would allow a company to develop the most effective communications message to consumers, maximizing attraction to its specific brand, product line, store, or department within the store.

Messaging and positioning choice modeling is recommended when the primary research objective is to obtain information that would allow a company to develop the most effective communications message to consumers, maximizing attraction to its specific brand, product line, store, or department within the store.

Component objectives of messaging and positioning choice modeling typically include:

• Isolate the most effective benefit statement or statements.
• Identify the most effective supporting statements.
• Determine the most effective or optimal combinations of benefit and supporting statements.
• Identify possible alternative supporting statements and measure their potential impact on the baseline (usually the optimal combination).

There are several general key benefits to choice modeling:

• Choice modeling defines product appeal in terms of choice behavior using a realistic setting. This allows management to know which items drive behavior and how strong and important they are.
• Choice modeling shows the power each product has in causing or driving the decision to buy. These models are causal—designed specifically to answer the question, “If I do this, then what would customers do?”
• Choice modeling simulates the future and allows forecasting and predicting of how customers and markets will react to specific decisions by management, allowing management to mix and match decision drivers to obtain optimum results before they commit to them.

Choice Survey Tasks

While participating in a survey, respondents see a number of different choice scenarios, or tasks. The information presented to respondents in the choice task scenarios places them in a situation where a choice or tradeoff must be made between several alternatives. For example, respondents are shown a total of 15 tasks with three alternatives per task. Each alternative is described in detail, and the respondent reacts to the scenario in a way that would reflect most closely an actual decision the respondent would have to make or would be involved in making.

Each alternative includes one or two core benefit statements and three to six supporting statements. Respondents rate their preference among three alternatives, each with a different combination of core benefit and supporting statements. Each respondent then selects the combination of core benefit and supporting statements that is most appealing.

The current paper illustrates, by two examples, how to optimize messaging and positioning with choice modeling. The primary objective of the first example is to obtain information that would allow a company to develop the most-effective ad campaign to not only maintain its old customers but attract new customers. Table 1 demonstrates a choice survey task, in which respondents are presented with a question: “Which of the following restaurants would you select based solely on the descriptions of the following characteristics?” Table 2 lists four core benefit statements and 17 supporting statements in the hypothetical restaurant study.

Table 1: Example Restaurant Choice Scenario

Table 2: Core Benefit Statements and Supporting Statements for a Restaurant Company

Positioning

Supporting Messages

The second example is to help a software company to launch the best claims for future development and brand positioning. Table 3 involves five core benefit statements and 19 supporting statements.

Table 3: Core Benefit Statements and Supporting Statements for a Software Company

Core Benefit Statements

Supporting Statements

Choice Modeling Experimental Design

A choice modeling research design involves selecting positioning statements in the choice tasks, which start with core benefit statements and are supplemented by supporting statements. An experimental design is created that assures measurement of the relative value of each core benefit message, the relative value of supporting messages, and selected interaction effects between core benefit and supporting statements.

Simulating Optimal Combinations

After developing the choice model, all possible combinations of core benefit and supporting statements are simulated. The optimal combination may or may not be the most desirable or operationally feasible combination for the company. In this case, the question becomes which combinations are (1) the most optimal and (2) the most feasible (i.e., in line with the company’s strategic goals and objectives)?

The optimal combinations sometimes include redundancies (e.g., multiple statements on food or customer service), which would make for awkward marketing messages. It may be desirable in some instances not to include every supporting statement, especially if it has already been emphasized by a previous supporting statement. For this reason, each particular combination relative to the optimal combination is indexed in order to answer the question: “How much is lost, in comparison to the optimal combination, by substituting different individual supporting statements?”

Combinations Charts

After simulating thousands of simulations, the top 100 simulations based on purchase interest can be selected. Table 4 (below) shows an example of a Combinations Chart. This type of chart is intended to uncover the messaging and positioning pillars that successfully combine to attract customers to restaurants or to the software brand.

Table 4: Combinations Chart for Top-Ten Positioning Strategies

The benefit statements in blue are on the left side of the chart alongside the supporting statements in green on the right. The chart is meant to be read across (i.e., from left to right) starting at the top row. Each row depicts either one or two benefit statements along with three, four, five, or six supporting statements, rank ordered from 1-100 (out of hundreds of thousands, or even millions, of possible combinations).

Each combination’s index, which is relative to the top combination, is presented to the far right. An index of 90, for example, indicates a preference of 90% of that which is obtainable with the optimal combination. By scrolling through the combination of benefit(s) and supporting statements, we are able to see a pattern emerge, whereby some statements appear essential to every top combination. In this chart, Benefit 2 is the critical lead benefit statement, while three supporting statements emerge as the most important: Support 5, Support 19, and Support 15.

In the analytical software example, the top positioning might be:

As shown in the examples in this paper, in some cases when clients are uncertain about what are the best messages to deliver to their customers, optimizing messaging and positioning choice modeling can help to identify the most effective or the most feasible benefits to define the brand and win business. When Decision Analyst has conducted projects to optimize messaging and positioning using choice modeling, the findings have proven to be extremely helpful in defining powerful brands.