CBC market simulator

The market simulator (also known as the choice simulator) transforms conjoint utilities into actionable insights by simulating market choices between different product alternatives. It is often considered the most crucial decision-making tool in a conjoint project.

By introducing products defined by the attributes and levels in your study, the simulator generates projections of the percentage of respondents likely to choose each product — or "None," if the none option was included. Think of it as a "voting machine" that simulates how your market might vote for each configured product. Results, or "shares of preference," always sum to 100%.

Under favorable conditions such as mature markets with equal information and distribution, market simulators can report results that closely match long-range equilibrium market shares. However, conjoint analysis cannot account for many real-world factors that shape market shares, such as time on market, distribution, out-of-stock conditions, advertising, sales force effectiveness, and awareness. Predictions also assume all relevant attributes influencing share have been measured.

Share of preference results should generally not be interpreted as market shares. Instead, treat them as relative indications of preference, reflecting how respondents might rank products based on the attributes and levels in the study.

The simulator starts with two empty products, allowing you to quickly compare two alternatives. At least one product is required to run a simulation.

Select a level from each attribute for each product. Once all attributes are filled in for the first product, the simulation runs automatically and updates with each additional product added.

It is generally best practice to specify a number of products that closely resembles the number of alternatives per task in your CBC questionnaire.

The first step is typically to define a base case scenario reflecting a current or future market: your brand versus the relevant competition. If there is no relevant competition, or your study was designed to model only your product, the base case may consist of a single product versus the "None" alternative.

Once the base case is set, add additional scenarios using the scenario selector in the top right corner to explore different market configurations and compare product alternatives.

The scenarios fly out is open showing a few scenarios have been created.

Votes are split proportionally across product alternatives for each respondent based on their likelihood of choice. For example, if two products are equally desirable, a respondent's vote is split 50/50.

Note: The share of preference model uses the logit equation. Product utilities are exponentiated and shares are normalized to sum to 100%.

You can optionally display 95% confidence intervals. Assuming your respondents are representative of the population, you can be 95% confident that the true population share of preference falls within this interval.

If confidence intervals for two product alternatives do not overlap, you can be at least 95% confident that one is preferred to the other.

Segmenting splits simulation results into groups based on responses or values from a question or variable in your survey. For example, segmenting by respondent location to compare North America versus Europe. Respondents without a value for the selected variable are excluded from results, and empty segment groups are not shown.

For quantitative attributes such as speed, price, or weight, you can configure them for use in interpolation, simulating values between the levels shown to respondents in the CBC questionnaire. Click the settings icon next to the simulator tab to access these settings.

Marking an attribute as continuous allows you to assign numeric values to each level, enabling more intuitive product specification and linear interpolation or extrapolation between attribute levels.

In the attribute settings area, assign a numeric value to each level. For example, 30, 40, and 50 for a weight attribute.

In the attribute settings overlay, weight has been marked as continuous and values of 30, 40, and 50 have been added to the table opposite the corresponding weight.

When configuring products, you can then select a predefined value or enter a custom numeric value within the range, or outside it for extrapolation. Note that extrapolated values can be unreliable since respondents did not evaluate them in the questionnaire.

A product in the simulator configured with a weight of 46.

Linear interpolation is applied per respondent on part-worth utilities prior to simulating shares of preference. For example, if level 1 of weight is 30 kilos and level 2 is 40 kilos, specifying 33 kilos interpolates a utility 30% of the way between the two. For respondent 1, if the utility for 30 kilos is 1 and the utility for 40 kilos is 0, the simulated utility for 33 kilos is 0.7.

Two files are available for download:

Scenario: An Excel file containing the configured products, preference scores, standard error, and upper and lower 95% confidence intervals. Scores are rescaled.
Charts: A PNG export of the scenario chart.

Discover Help Guide

CBC market simulator

Introduction

A note on interpreting results

Building products

Simulation method

Confidence intervals

Segmenting

Simulator settings

Attribute configuration

Continuous attributes

Technical details

Downloads