Value proposition testing with conjoint analysis

How to establish a robust and reliable testing environment

Every great innovation starts with discovery and exploration. But when you start exploring a new business idea, whether being a start-up or an incumbent, you usually find yourself in a space of very high uncertainty. You don’t know if your idea will work. So you start mapping out the idea; continuously designing and redesigning the idea until you have something that looks like a viable value proposition. But either of these things will make your idea a certain success.

That makes your chances of succeeding with your new idea less than 30% if you do not take the appropriate measures to beat the statistics. To do so, you must close the knowledge gap between strategy and experimentation in order to decrease the risk of wasting time, money and other resources on product initiatives with no market fit. In other words, once you have explored your idea and designed your value proposition, you must experiment and test assumptions that are key to the success of your value proposition.

The following illustration sums up our approach to value proposition design.

Image for post
Image for post
Research approach

Naturally, the commercial success rate depends on what we end up testing, and thus the exploration and design phase is important. However, it also depends largely on how we test the idea.

That said, testing always comes with several challenges such as:

  • How do I turn jobs, pains and gains into something testable in a reliable way?
  • How do I create a reliable testing environment that yields results that reflect the real world?
  • How do close the knowledge gap between strategy and experimentation, translating data into successful decisions?

In the following, we will show how one statistical tool excels at solving these exact challenges, and even outmatch other methods by far, when applied in the context of a robust process. This tool is called conjoint analysis.

Conjoint analysis in short
Conjoint analysis is a complex type of market experiment simulating a real-world purchasing situation. Respondents are presented to several scenarios with varying product alternatives, typically at different price levels. During the experiment, respondents evaluate the different alternatives and related attributes, and thus indirectly reveal their preferences for, or perceived value of, attributes when choosing between the alternatives.
Through statistical modelling, insights like optimal price, market potential, price sensitivity, preference for each feature and much more can be derived, which then can constitute the foundation of e.g. go-to-market strategies.

In this case, we will walk you through a real-life business case we solved with Volkswagen.

Starting with the problem

Generally, one should always start with the problem before progressing towards the solution to the problem. This approach allows for steering clear of wrong assumptions by adjusting and clarifying hypotheses on an on-going basis.

The basis of the study was quite simple. Volkswagen operates in a mature market that is currently undergoing a lot of changes in terms of technology, consumer preferences and regulation.

In that kind of market, retention is key. Maintenance plans is an excellent tool to keep customers in the business, but Volkswagen was underperforming on penetration of maintenance plans.

With that in mind, our point of departure was:

  • How can Volkswagen increase penetration on maintenance plans significantly, and where should we look for the solution?

Exploring and designing

To make sure the results were unbiased, we first conducted customer discovery interviews. This allowed us to establish a customer profile consisting of jobs, pains and gains before agreeing on the ranking of jobs, pains and gains.

After conducting and analyzing the interviews, we could agree on where to look for the solution to the problem. Thus, the research was split into 1) identifying critical and underperforming steps in the pre-purchase process, and 2) measuring the value the maintenance plan itself captures by applying conjoint analysis.

The following illustration sums up where we would search for reasons for Volkswagen underperforming on penetration of maintenance plans.

Image for post
Image for post

Testing, experimenting and learning

As shown above, our natural starting point for the pre-purchase process was the sales funnel. Based on open-ended interviews, we were able to pick out specific steps in the sales funnel that customers find critical and test them. For the purpose of this article, we will focus on the conjoint part of the study. That is, the solution composition part.

To measure what value the maintenance plan captures, it was necessary to break the plan down into individual components, or attributes, and relevant levels, which are shown in the below illustration. Naturally, these attributes and related levels were determined strictly based on the interviews.

Image for post
Image for post
Maintenance plan attributes and related levels

With the conjoint setup in place, customers were invited to participate in an online survey combining a regular quantitative study and a conjoint market experiment module.

With the untreated conjoint data at hand, we could start analyzing. Among the most interesting insights were what customers actually find most valuable when deciding what maintenance plan to purchase. Loan car, which is not currently included without extra costs, proved to be the most valuable attribute after price. It was even possible to calculate exactly how much customers were willing to pay to include the loan car. Similarly, willingness-to-pay and preferences for the other attributes and related levels could be derived, and a lot of insights needed for optimizing the maintenance plan composition were present.

Another very useful insight that conjoint analysis can provide is an optimal price. Based on attribute preferences, several customized market scenarios can be established taking a given maintenance plan and competition into account. From there on, the optimal price for all possible combinations of maintenance plans can be calculated with related market shares, revenue and even profits if cost structures are available.

Testing environment: Applying the conjoint results

As one of conjoint analysis’ strengths as a method is simulating a real purchasing situation, we can trust the results to reflect customers’ true preferences. This is partly due to customers evaluating alternatives and thereby indirectly stating preferences in stead of directly stating them e.g. verbally. This line of thought is equally important when interviewing.

It does, however, presuppose that the attributes and related levels are the most relevant in the eyes of the customer. Note that this is why carefully exploring and designing is so important. When done right, however, results are extremely useful and has the potential to provide insights that can be applied directly in strategic decisions.

In conclusion, conjoint analysis has the power to establish a reliable testing environment that, in a business strategy and product management context, provide clear and understandable results in a relatively inexpensive way, and thus reduce the risk of failure.

Want to know more about experimenting and testing? Check out the latest publication in the Strategyzer book series “Testing Business Ideas”

Written by

We are ag analytics. A hybrid tech analytics consultancy combining the best in strategic thinking, data science and preparatory technology. www.aganalytics.dk

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store