How to enforce non-positive price coefficients on Conjointly?
The option to enforce non-positive price coefficients for Generic Conjoint and Brand-Price Trade-Off (BPTO) is available, but we do not recommend enabling this option unless you have extremely strong reasons to believe in the assumptions about preferences for lower prices.
Generally, the assumption about consumers’ preferences for lower prices holds true especially for FMCG products. However, there are at times when consumers would prefer higher prices instead, with one of the possible reasons being price-quality inference (i.e. higher price indicates better quality). This results in a positive price coefficients, and will have the following consequences:
- Marginal Willingness To Pay (MWTP) from Generic Conjoint will not be available
- Preference share simulator will show greater preference share when the product concept has higher price and show positive price elasticity of demand (PED)
If you have a strong reason to believe that the consumers will always prefer lower prices for your products/services, and would like to avoid having positive price coefficients, you can take the following steps to enforce a non-positive price coefficients:
- Go to the Advanced settings tab of your experiment and look for the Analytics options module
- Check the option for “Enforce non-positive price coefficients”
- Click at the bottom of the page
- Click and your report will be recalculated to apply the changes
By enabling this option, Conjointly will rescale the price coefficients to be non-positive. Hence, the MWTP output from Generic Conjoint will be available, while the preference share simulator will not show greater share for higher prices and positive PED. However, you should be cautious in interpreting these results because this option will likely underestimate the impact of price.
Comparison of outputs before and after enforcing the non-positive price coefficients:
Before enforcing non-positive price coefficients
- MTWP output is not available in Generic Conjoint because respondents do not necessarily prefer lower price
- The preference share simulation output shows greater preference share with a higher price, for example: A market simulation of 3 product concepts with the same configuration but different pricing ($24,000, $28,000, and $32,000) shows the following results:
- Product concept 1 with price $24,000 takes ~19% of share
- Product concept 2 with price $28,000 takes ~18% of share
- Product concept 3 with price $32,000 takes ~61% of share
After enforcing non-positive price coefficients
- MTWP output is available in Generic Conjoint after the readjustment of price coefficients
- The preference share simulation output minimises the impact of price on preference share, for example: The same market simulation shows a more similar distribution of preference share for each product concept:
- Product concept 1 with price $24,000 takes ~34% of share
- Product concept 2 with price $28,000 takes ~33% of share
- Product concept 3 with price $32,000 takes ~32% of share