Improving Building, Home Improvement, Lawn and Farm Products, Pricing with Conjoint Research
Why Use Conjoint Methodologies?
“On a 1-10 scale, please rate the importance of the following product attributes”—this is a typical question used in market research to gauge the hierarchy of product features for a new concept. The problem … most consumers don’t actually use the entire 1-10 scale. Really the question should read “On a 6-10 scale, please rate the importance of the following product attributes” because most people only use the upper half of the scale. Consequently, it gets a little more difficult for manufacturers to confidently produce a product with features that will accurately resonate with the consumer–let alone at a price that will capture market share.
Gauging which product features and prices will in turn move the production needle is a difficult task that can easily get bogged down by the difference between an 8.6 and an 8.7 on that 10-point scale. So, what is a manufacturer to do? This is where conjoint analysis can come in to help make things clearer. Conjoint expert Keith Chrzan from Sawtooth Software explains the methodology well,
“It is a technique that helps cut through the minimal difference that we find among attributes in our 1-10 rating scales, and instead force the respondent to make product feature tradeoffs in a simulated purchase exercise. Those tradeoffs help us figure out how much unique value each feature adds to the product.”
There are three common conjoint methods that can be used depending on the goals and needs of a project: choice based conjoint, adaptive choice based conjoint, and menu based conjoint. A brief overview of all three is given below. When taking on a conjoint project you should research further the questions and problems that each method can solve—each can deliver a wealth of analytical direction.
All three of these conjoint methods simulate a buying experience for a respondent. This purchase simulation is a series of screens that a respondent walks through on an online survey. Each screen displays around five product attributes for three to five products. Those products will all have the same primary attributes—in our example they are Coating, Warranty, Brand, Material and Price—but each product can have different secondary attributes or levels. In our example the secondary attributes to Price would be $55, $120 and $180.
Choice based conjoint (CBC) is the most used of the conjoint methods, and it also happens to be the simplest method. In this shopping simulation, depending on the study’s sample size, respondents will see 10-20 screens like our example here—each respondent then picks the product they would prefer out of the 3 presented. Once all the 10-20 screens have been completed for all participants, the data can then be analyzed to formulate a hierarchy of attribute effectiveness. The sample size requirement is large for this type of conjoint, but because it is the simplest of the conjoint methods, it typically is the most economical method.
Adaptive choice based conjoint (ACBC) is a version of CBC, but is set up in a way that as a respondent is working through the simulation exercise the simulator learns what their preferences typically are and adapts the simulation as the respondent proceeds through subsequent screens. The subsequent screens will then focus on those secondary product attributes that are most preferred by the respondent. Data is then extrapolated similarly as CBC. A bonus for this type of conjoint is that it does not require a large sample size, however the programming and analysis costs are typically higher.
Menu based conjoint (MBC) is a version of conjoint that seeks the respondent’s preferences for products up front as opposed to ACBC which learns preferences along the way. Much like ordering off a menu, the respondent picks the types of features (primary and secondary) that are important to them. Once this menu of product attributes has been viewed, the respondent then proceeds to a shopping simulation much like the CBC example above. This type of conjoint is great for complex projects, but a downside to this conjoint is that it requires a very large sample size which will increase the costs of a project.
If the needs are appropriate, conjoint methods can give a better sense of product feature importance and priority, it will also give price elasticity of the various product feature scenarios run in the simulations. When utilizing conjoint, you can be more confident in what product features, and at what price, capture the most market share from your competitors.
Contact us for a consultation on your needs and how research designed around your objectives can get you answers.
Written by: Taylor Pence, firstname.lastname@example.org