February 2025 | Advanced Analytics

Three Advanced Analytics Applications
And What They Can Do for Your Brands

At Decision Analyst, we believe solving major strategic issues requires deep thought, rigorous analytics, and creative vision. Decision Analyst’s Advanced Analytics Team applies scientific thinking and scientific methods to the solution of strategic business and marketing issues.

When incorporating advanced analytic techniques in research projects, the central focus is the scientific determination of “cause and effect” through rigorous research designs and carefully controlled experiments.

Here are three applications of Advanced Analytics techniques to help resolve major marketing issues.

Choice Modeling Optimization

The term “choice modeling” refers to a family of trade-off and conjoint techniques used to optimize sets or combinations of marketing variables. Carefully controlled experiments are designed in which consumers are simply asked to choose how many of each product they would buy, given predetermined sets of realistic conditions. Each scenario has different variables (such as different brands, different prices, different messages, different package designs, and different positionings), and respondents are asked to choose which brand/product they would purchase. From the choices consumers make, predictive equations are derived to estimate the results of every combination of marketing variables.

Example Choice Model Question

Choice modeling applications:

• Optimize pricing strategies
• Optimize bundles of product features or service possibilities
• Optimize brand strategy
• Maximize sales (and/or profits) of a product line
• Optimize strategic positioning and advertising messages
• Optimize the appeal of package design elements
• Optimize new product concepts
• Optimize shelf sets, package sizes, number of facings
• Optimize the product mix in a vending machine

As this list of applications suggests, choice modeling can be successfully applied to a wide array of marketing optimization tasks.

DecisionSimulator™

Once choice modeling experiments are completed, equations are derived to build a DecisionSimulator™. The simulator allows research analysts or brand managers to play “what if” games. By changing the marketing variables’ inputs, the brand manager can determine how various marketing actions will affect brand preference or brand market share.

Pricing Optimization

Pricing is one variable in a complex set of variables that make up a brand’s overall pricing strategy. These variables include product features and quality levels, pace of innovation, speed of delivery, strategic positioning, advertising messaging, advertising spending levels, package design, and distribution channels. The central question is: What combination of variables enables a brand to consistently maximize its pricing vis-à-vis the pricing of competitive products and services?

Decision Analyst uses an array of research techniques and analytic methods to optimize pricing, including:

• Hierarchical Bayes Choice Modeling Within Competitive Set
• Joint Stated-Revealed Preference (JSRP) Modeling
• Artificial Intelligence and Machine-Learning Models
• Gabor-Granger Technique

Typically, a DecisionSimulator™ is one of the deliverables for pricing optimization studies, so that analysts and brand managers can explore many different pricing scenarios and react quickly to the pricing actions of competitors.

Market Segmentation

Market segmentation is a process of subdividing a large market into a number of segments so that a brand can focus on one or two segments rather than the whole market. By concentrating marketing efforts on one or two segments, a brand’s chances of success are greatly improved.

There are many different statistical techniques that can be used to segment a population, including artificial intelligence and machine learning, cluster analysis, factor analysis, discriminant analysis, and latent class segmentation. A best practice is to segment a dataset using multiple statistical techniques to determine the method that provides the very best segmentation solution. The size and relative potential of each market segment is calculated along with the positionings and appeals that should be employed to reach each segment. Market Segmentation answers five basic strategic questions:

• How is the overall market structured and organized?
• What market segment or segments should be targeted?
• What is a brand’s optimal strategic positioning to reach its target segment(s)?
• What advertising themes and messages should reinforce the strategic positioning?
• What new product opportunities exist across the market segments?

Segmentation StrategicImpact™

To achieve the greatest strategic impact, the following phases are recommended for major segmentation projects:

Qualitative Insights

Market Segmentation studies benefit greatly from good, upfront qualitative research to identify important psychographic variables and behaviors, to inform the design of the segmentation questionnaire. Qualitative research helps identify the consumer's language and vocabulary, reveal the ranges of variables, and guides quantitative analyses during the segmentation itself. Qualitative research is widely used on the backend to develop personas for the market segments identified during the study. Think of a persona as a "personality" to represent or typify a market segment. These personas help all stakeholders better understand and apply the segmentation results to improve marketing decisions.

Jerry W. Thomas

CEO

Email Jerry

Bonnie Janzen

President

Email Bonnie

Felicia Rogers

Corporate Executive Vice President

Email Felicia

Elizabeth Horn, Ph.D.

SVP, Advanced Analytics

Email Beth

Clay Dethloff

SVP, Qualitative Research & Innovation

Email Clay