Conjoint preference share simulator
Conjoint.ly allows you to simulate shares of preference for different market offerings. Specifically, you need to describe offerings that are available on the market in the language of the attributes and levels that you chose to include in the study, and the system will estimate the percentages of preferences for these offerings. Please keep in mind that:
If you would like to test out a new product, add it to the list of existing concepts.
If two or more offerings share exactly the same levels, input each offering separately. For example, if there are two brands each offering identical products, put in both brands in two separate lines.
The market shares are only approximate and may be affected by constrained availability of some products (e.g., not available in all geographies), shelf space (in case of FMCG).
For example, you can specify these three offerings to see approximate shares of preference:
|Brand||Monthly fee||Mobile data inclusion||International calls inclusion||SMS inclusion||Share of preference|
|Telstra||$49.00||500MB||0 min||300 messages||30%|
|Vodafone||$39.00||10GB||90 min||Unlimited text||20%|
|Optus||$45.00||Unlimited||300 min||Unlimited text||25%|
|None of the above||25%|
Models for calculating market shares
There are two models for calculating market shares:
Share of preference model, which is appropriate for low-risk or frequently purchased products: FMCG, software, etc. This model is applicable in the vast majority of applications.
First choice model, which is suitable for high-risk or seldom purchased products: education, life insurance, pension plans, etc.
In most cases, both simulation models produce substantially the same results. However, at times, simulations results will differ greatly from actual market shares, hence adjustments may be needed.
Automated preference share simulations for what-if scenarios
Brand Specific Conjoint experiments provide outputs allowing estimation of; market overview, overall profile, segmentation and interactive simulations. Preference shares, revenue projections and price elasticity can be estimated and charted using Conjoint.ly simulators.
In the below scenario a baseline for four car brands is being shown for preference shares and revenue projections.
Attributes and levels that have been set in the experiment design appear in the simulations screen allowing for charting all possible scenarios. Each scenario can be manually renamed and allows for advanced options such as; segmentation, manual availability adjustment of observed volume share, and adjustment of relative availability (adjustment factor).
|Make and model||Drive-away price||Transmission||Engine type|
Preference shares and revenue projections (assuming 1,00 units offered) can be simulated to assist in adding or modifying product features.
Price elasticity of demand in two clicks
Price elasticity of demand can be calculated for multiple scenarios by clicking on two points of a preference share simulation.
In this scenario, when clicking on both the $23,000 and $25,000 price points for Ladina Klubnika a message will be displayed stating that the price elasticity of demand is elastic.
Price elasticity of demand between $23,000 and $25,000 is -4.6.
Demand is elastic (i.e., an increase in price by 1% leads to more than 1% drop in volume).
Apply custom segmentation to simulations by clicking the “show advanced settings” button and select the required segment(s).
New product launch simulations
New product launches (NPDs) can be simulated to see redistribution of preference and revenue shares.
Excel simulator example
The segregate function creates a variable based on people who would buy certain concepts within each scenario. It allows you to then segment your report to view the preferences of respondents who are willing to buy certain products. To enable the segregate function:
- Open the "Simulations" tab in your survey report.
- Click the "Show advanced settings" button at the bottom of your screen then click the "Segregate" button next to the "Enable segmentation based on preferences for concepts in this scenario" heading. A message will appear in the left panel once segmentation is complete.
- Open the "Segmentation" tab and click the "Add a new condition button.
- Click the "Segregation based on simulation" option.
- Select your desired Variable, Type of match, and Value from the respective drop-down menus before clicking the "Save and apply these segments" button in the bottom right corner. A message will appear in the left panel once segregation is complete.
- Open the "Compare segments" tab to view respondents' profile and preferences from your new segment.
Interpreting the Excel simulator
To interpret the Excel simulator, you can trace how it works from the hidden sheets labelled ‘SimXXX’. For example:
Sim1 contains individual preferences.
Sim2 translates the simulator parameters into 0s and 1s.
Sim3 is a matrix obtained by MMULT of Sim1 and Sim2.
Sim4 to Sim6 calculates the individual share of preference through the multinomial logit formula:
MaxDiff preference share simulator
Using the simulator for MaxDiff should be done with caution because the simulator answers the question “What % of time would a particular option be chosen?”. Choice-based questions (i.e. Conjoint or Claims Test) can accurately answer the simulator’s question because they are about choice among several options whereas MaxDiff asks both “best” and “worst” choices and uses the “worst” choice in calculating utilities as well. This means that an option that is occasionally named “worst” in a choice set will have a lower predicted preference share in MaxDiff than it would in conjoint (where we do not consider “worst” options), which can cause overestimation of preference shares for the topmost preferred options. For this reason, we suggest using simulations for choice-based exercises only as these are robust whereas MaxDiff simulations are only indicative.
To see the above features of Conjoint.ly in more detail you can start experimenting with the Conjoint.ly simulator using “Example experiment 2: Preferences in cars (brand-specific)” in your example experiment library.
Here are also some suggestions for further reading:
- What is conjoint analysis?
- Top tips for specifying attributes and levels in conjoint analysis
- Example report on preferences in ice-cream
Would you like to see more example conjoint reports? Log in to explore example reports