Conjoint analysis is a popular method that uncovers consumers’ preferences. In an automated manner and in only a few steps it mimics the tradeoffs people make in the real world when making choices. Let's see how you can implement an end-to-end Conjoint method using Nebu solutions.
Conjoint Analysis is a market research technique primarily used for measuring how valuable certain features of a product (or service) are to its users.
The respondents of a Conjoint survey are required to compare different products. The products are described with the same attributes but with different values, which makes combinations comparable. For instance two cups of coffee can be compared based on their caffeine contents.
There are multiple approaches to implement a Conjoint Survey.
The least complicated one is the rating-based version. In this implementation, only one product is displayed per page with all of its features. The respondent’s task is to rate (typically from 0 to 100) each presented product. The disadvantage with this implementation is that it doesn’t really force respondents to compare the products. They may enter the same value for each product and complete the questionnaire with minimal effort. Due to this, it is the least popular of Conjoint techniques.
The implementation we are focusing on in this article is Choice-based Conjoint. It is also the most common way of doing this kind of research.
The Choice-based Conjoint approach's main characteristics are that multiple products (typically 3-5) with the same attributes of different values are displayed per page. The respondent then has to select the most favorable one.
This approach is prevalent for two main reasons.
On the one hand, the survey is much shorter as not all feature combinations have to be rated separately. On the other hand, because the survey respondents are forced to make trade-offs (like choosing one feature over another), it simulates an actual, real-world buying scenario.
Not surprisingly, the first step when designing a study like this is to identify the products to compare. What attributes do they have and what values can be assigned to them? If the products have other attributes that we don’t want to include in the comparisons, it’s also important to inform the respondents about them.
After we have decided on attributes and their values, the next step is to create a questionnaire design. It involves combining the attribute values into actual products and deciding which ones to show on certain questionnaire pages.
Our aim here is to make the questionnaire as short as possible while keeping the study reliable by making the right comparisons.
Once the design is set up, the next step is to build the actual questionnaire. Of course, the implementation depends on the data collection tool. In my opinion, the best practice would be to create a general template that can dynamically build the questionnaire based on the design created in the previous step.
It’s also important to include the following information in the questionnaire’s dataset: which products with which features (attributes) were displayed, which one was chosen as most preferred.
Implementing a workflow like this can be a really challenging task if you have to build everything from scratch, as it requires quite some statistical and programming skills.
Fortunately, Nebu Data Suite has a ready-to-use implementation for Conjoint questionnaires that you can add to your projects with a relatively low effort. It consists of two main parts: an R script that is used for creating the design. It can be parameterized and run in Nebu Data Hub. The second part is a Nebu InterViewer questionnaire which can pick up the design and build its pages accordingly. You can check out a Nebu Conjoint demo questionnaire here. What is great about this solution is that it is ready after setting a few parameters but can also be easily customized if you need something more advanced.
Now let’s talk about what we can do with the dataset of a Conjoint survey.
Our aim is usually something like finding the most important attributes that weigh in the most, or the most preferred combination of features that make the perfect product. Usually we want to identify accurate behavioral patterns driving the customers so that we can validate current - or drive the future - business decisions.
For such purposes, we can use Logit Models.
A Logit Model is a statistical model with which the probability of certain outcomes of an event can be predicted. In the case of Conjoint analysis, this event can be for example, a customer buying a certain product with a certain feature set. From survey respondents' behavior, we can set up generalized predictions about how preferable a certain product would be on the market.
Please check out this dashboard dashboard that was created from the dataset of the previously mentioned Conjoint questionnaire about the possible outputs of a Logit Model analysis. If you are a registered Nebu Data Suite user you can go to this support article to learn how to implement a Conjoint study.
If you want to learn more about Nebu Data Suite and how it can help you automate processes and increase work efficiency submit a form to the right. Our expert will reach out soon to schedule a call!