Conjoint Analysis Example Simulator & Business Questions You Can Answer
You can use conjoint analysis to answer a lot of business questions. Most of them are related to product and price. We are going to use a simulator example to show you how to answer some of the most common business questions that come up.
This article will give a brief introduction to a few key areas. You can click on each section for an even more detailed video and write up on each.
Please note: conjoint analysis is a broad term. The concept applies anytime you have multiple stimuli that you ask the target audience to respond to. Discrete choice is the most powerful way to do conjoint analysis, but on this website favor the more general term and will use the terms interchangeably.
How Does Demand Change with Price?
In other words, how much should the company charge for it's products to get the most profit?
You may have heard the phrase: let's lower the price and make up for it on volume. This is usually a bad idea unless you know exactly how much volume you would need to make it up and the feasibility of reaching that target.
The simulator will help you discover the relationship between price and revenue/profit.
How Does Demand Change when Feature X Is Included?
Which features would be the most profitable in our products? Each candidate feature will have a different value (utility) among consumers. They, potentially, will also have a different cost associated with delivering its benefit. The conjoint simulator will help us find the sweet spot between demand and cost that results in the greatest profitability for the product or line.
How Are Units, Revenue and Profit Calculated in the Conjoint Simulator?
The statistical model predicts the probability that each options will be chosen. So how do we go from that to profit?
The math is simpler than you might guess.
How Do You Calculate Source of Volume in the Choice Simulator?
It's a common and important question: If we make this product better and increase sales, will the new sales come from our existing products or from competitors? Before finalizing a product strategy, you need a good understanding of where you are likely to be sourcing the new volume.
How Do You Add New Products into the Market?
Maybe you have a new-to-the-world feature or benefit that will be at the center of your new product. Or perhaps, you are planning on entering a new or adjacent category. Conjoint analysis will help you forecast how consumers will react to the new marketplace and provide a forecast of sales and profitability.
How Do You Calculate Willingness to Pay in the Conjoint Simulator?
You can use the conjoint simulator to identify how much consumers are willing to pay (WTP) for different product or service features. In the article we discuss a new method for calculating WTP that is better aligned with the needs of decision makers.
In contrast with the old method the new way allows for decision makers to understand how much additional price the product can command in the market with the feature change.
The method also provides a forecast for revenue and profit at the new price point.
Why Didn't My Choice Study Get Nice Price Curves?
uh oh...
![Conjoint Analysis Price-Revenue Curve [Terrible]](http://redanalytics.net/wp-content/uploads/2018/11/Conjoint-Analysis-Price-Revenue-Curve-Terrible.png "Conjoint Analysis Price-Revenue Curve [Terrible]")
Up until 2016. Price-Revenue curves 'always' suggested the highest price point would maximize revenue. We just weren't picking up price sensitivity properly. First, we thought it was an artifact of the experiment where people didn't really have to part with their money for the choice made. We also considered different model and parameter specifications. It was hopeless.
Then finally the answer came out of academia. Respondent quality matters - A LOT.
The analytic procedure is complicated, but by identifying and removing information-poor respondents the Price-Revenue curves fix themselves.
Conjoint Simulator Assumptions Needed
There are several areas where the default is much worse than using a smart guess about assumptions. You'll need information about product/brand awareness and distributions and cost information.
Keep reading to learn about the 3 different types of costs and a useful structure to keep the costs well organized.
References
Batsell, R.R. & Louviere, J.J. Market Lett (1991) Experimental Analysis of Choice 2: 199. https://doi.org/10.1007/BF02404072
C Breidert, M Hahsler, T Reutterer (2006) A review of methods for measuring willingness-to-pay Innovative Marketing
Dotson, Jeffrey and Brazell, Jeff D. and Howell, John R. and Lenk, Peter and Otter, Thomas and MacEachern, Steven N. and Allenby, Greg M., A Probit Model with Structured Covariance for Similarity Effects and Source of Volume Calculations (April 1, 2015).
Green, P., & Srinivasan, V. (1990). Conjoint Analysis in Marketing: New Developments with Implications for Research and Practice. Journal of Marketing, 54(4), 3-19. doi:10.2307/1251756
Rao, V. (1984). Pricing Research in Marketing: The State of the Art. The Journal of Business, 57(1), S39-S60. Retrieved from http://www.jstor.org/stable/2352922
