We often tell our clients they should pay more attention to how their customers behave rather than what the customers say. We don’t mean to imply that clients should ignore customer feedback; but customers do not always act the way they say they will. For example, they will often say price is the most important thing, and then decide what to purchase and where to purchase it based on other criteria. One helpful technique for understanding real customer behavior is conjoint analysis, which is a market research tool that measures customer trade-offs.
Conjoint analysis is sometimes also called discrete choice analysis. In a conjoint, customers or prospective customers (respondents) are typically shown three to five product or service alternatives on a screen, and they are asked to pick just one. Each of the products will have a list of attributes, such as size, capacity, durability, price or other important variables. The respondents select their preferred product but do not indicate why. They are asked to repeat this selection process several times (often 12 to 30), and each time the mix of product attributes changes. With enough respondents answering enough questions, the conjoint can statistically determine which attributes are most important, and by how much.
Our clients’ transaction histories are usually rich data sets they can mine to understand their customer behaviors. By measuring how customers react to specific price changes, we can determine which customers are elastic and which are not, and on which products. We can also study wins and losses with a variety of variables to understand how much customers value their current products and incumbent suppliers, whether they will trade larger orders for lower prices, how important service convenience is, etc. However, all of those are looking through a rear-view lens. When past transactions may not be a good indicator of how customers might react to a change, or you are contemplating something completely new, conjoint analysis can help.
Let’s consider the example of a new product you have in development. You want to know how much demand there might be for that product, and at what price. The more revolutionary that new product is, as opposed to a me-too version of existing products on the market, the less you can rely on indicators of your past transactions. You can learn a lot about what will be important by getting a group of customers or prospects to make a series of choices between the new product with its attributes and prices versus existing products with their own attributes and prices. You can also use this technique to determine if the prices customers might be willing to pay for your new product will be worth the investment required to develop it.
A second situation that would be appropriate for a conjoint analysis is to learn the values of new features you are considering offering that are not available in competitor products. For example, perhaps you are considering offering a lifetime warranty and all competitors offer a three-year warranty. That extra warranty is worth something, but how much? One way to find out is to introduce the longer warranty and increase your price, but if you raise your price too much, you may lose sales and profits. Increase it too little and the extra revenue may not cover the cost of providing and administering the warranty. A conjoint analysis is a good way to test the idea of the lifetime warranty without actually delivering it. You would only need to offer it if the conjoint shows customers value it sufficiently to provide an appropriate ROI.
Another example where discrete choice analysis can be useful is in understanding how B2B customers value different service levels of competing product sellers. In a conjoint, you can test how often customers make trade-offs to obtain things like online reporting, payment terms and payment options, customized packaging, etc. Knowing how important each attribute is and how much customers will pay for them will help you plan how to communicate the areas in which you add value, and to price them more appropriately.
We (most people) often do things counter to what we say we will do. We don’t always eat as healthy as we plan to, or exercise as much as we promise. The psychological reasons for that are similar to the reasons customers say they care most about prices but buy things for different reasons. And that is why we tell our clients they need to understand how customers are likely to behave, not just listen to what they say. Analyzing customer purchase behavior within their own transactions is an important part of that; however, in many situations, their understanding will be substantially improved by utilizing conjoint analysis.
Interested in learning more about how conjoint analysis can help? Contact us.
Leave a Reply
Want to join the discussion?Feel free to contribute!