What is choice modeling?

Your customers can tell you what they like and what they’ll buy, but they can’t always explain why they choose one brand over another. Unless they are marketers themselves, they may not fully understand the role of price, brand image, packaging, brand name, promotions, and advertising in their decision to purchase. 

Choice modeling, a type of preference structure modeling, is a powerful tool for understanding what drives customer interest and purchase decisions. It’s considered to be the most scientifically robust way to discover and understand how customers make choices.

Let’s take a closer look at choice modeling and how it can fit into your marketing strategy. 

Choice modeling is an analytical method that is used to simulate consumer shopping behavior.

Research participants are unaware of what is being measured as they are presented with visual choices with marketing variables such as advertising, pricing, packaging, features, etc. Participants are asked to make trade-offs among the provided options, ultimately choosing what they value most from those options.

Inferences drawn from participant decisions are used to predict the likelihood of a customer choosing one product or feature over another.

The data provides deeper insights into what is important to your target market, allowing you to make insightful, data-driven business decisions for various dilemmas, including:

The most significant advantage of choice modeling is that it provides deeper insight into your target market’s values. Other advantages include:

As with any research method, there are limitations to be considered. These limitations include:

Choice modeling says that individuals make decisions based on weighing the utility of each alternative, choosing the option with the highest utility. This is accomplished using the logistic statistical model to determine the probability of future events.

Illustrative benchmark:

On the SurveyMonkey LaunchPad Price Optimization solution, price-based choice data from a Van Westendorp study is plotted against an acceptable price range to surface a single optimal price point, for example $10 within an $8–$12 acceptable range.

Choice modeling studies aimed at "price setting for profitability" follow the same principle: the goal isn't just to rank preferences, but to convert those trade-offs into one specific, defensible number a pricing or product team can act on.

There are several types of choice modeling analysis. The type you use depends on your technical resources and the kind of data and insights you are seeking.

R is a free, open-source programming language created by statisticians for working with data. R-Language is frequently used to analyze large datasets with complex variables. R can handle both discrete (nominal or ordinal) and probabilistic variables. It runs on a wide variety of UNIX platforms, Windows, and Mac OS. R can be used for statistical analysis and visualization of your SurveyMonkey data. R is known for being difficult to learn for those with limited experience in programming.

Another type of choice modeling is conjoint analysis, also known as trade-off analysis. Conjoint analysis is based on the concept that any offering from a company can be broken down into a set of attributes that impact a customer’s perceived value of the offering.

Use conjoint analysis to determine the most influential attributes on a survey participant’s decision to purchase. 

Your survey structure for conjoint analysis should ask participants to rank the importance of specific attributes or to choose between different combinations of features and prices. 

During analysis, a value is assigned to each attribute. The data can then be used to decide the combination of features that will be most attractive to customers and at what price they are willing to make a purchase.

Yet another method of determining the probability that a consumer will choose a particular alternative is discrete choice modeling. This is best for product categories that see one purchase used over a long period of time or products that have many features, such as smartphones.

In discrete choice modeling, both current and potential customers are asked to view a realistic scenario that includes all of the competing products in the marketplace. They are then presented with varying combinations of marketing strategies and asked which product they would purchase based on that marketing. 

Volumetric choice modeling is common for businesses in product categories that experience multiple product purchases in short amounts of time and where repeat purchase volume is important. In this type of modeling, current and potential customers are provided with a realistic shopping scenario that includes all of the competing products in the particular marketplace. They are asked to indicate how many of each product they would buy. This reveals the role and importance of marketing variables in a situation where varying quantities of multiple brands can be purchased.

Choice modeling is a market research method that predicts real-world purchasing behavior by asking people to make trade-offs between product attributes, such as price, features, and packaging, rather than rating each attribute in isolation.

A choice model is the statistical model built from those trade-off responses. It estimates the utility, or relative value, that each attribute holds for respondents, then uses those utility scores to predict which option a customer is most likely to choose.

The choice modelling method covers the four main analytical approaches: R-Language modeling, conjoint analysis, discrete choice modeling, and volumetric choice modeling, plus related best-worst scaling methods like MaxDiff. Each method presents respondents with trade-off scenarios and uses their choices to infer what they value most.

Choice modeling market research gives teams a scientifically robust, data-driven answer to "why" customers choose one option over another, which stated-preference surveys alone can't reliably capture. That makes it useful for price setting, feature bundling, product positioning, and concept testing.

Choice modeling effectively determines what's important to your customers and potential customers when making purchase decisions. Start using choice modeling today to test product features for implicit value, the effectiveness of marketing campaigns, set pricing structures, and more.

SurveyMonkey has a variety of market research services available, including product optimization, price sensitivity analysis, and survey design. Explore the SurveyMonkey LaunchPad MaxDiff solution to rank features, messages, and claims by preference without building the statistical model yourself.

If you'd rather hand the legwork to someone else, SurveyMonkey Market Research Services can design, field, and analyze a custom study for you, from questionnaire design through the final report.