What is the customer willing to pay for the product? Which product features are important for selling the product? Which features should the product development team prioritize? Which features are more important, and which are less important?
Have you ever asked yourself some of these questions and wanted to know the answer in an objective way? If so, you should consider implementing conjoint analysis into your market research.
In this article, you will learn what conjoint analysis is, how to design and execute it, and read examples of its implementation within product teams.
Table of contents
- What is conjoint analysis?
- Key elements of conjoint analysis
- Conjoint analysis example
- How to design and execute a conjoint analysis study
- Conjoint analysis case studies
- Advantages and disadvantages of conjoint analysis
- Tools to design and execute conjoint analysis
What is conjoint analysis?
Conjoint analysis is a statistical method often used by product managers to conduct market research and evaluate how customers value different product attributes.
For product managers, it’s important to know which attributes of the product increase the perceived value for the customers the most. This way you can focus on the most valuable features first and gain higher returns on investments in the development of the product.
Conjoint Analysis is one of the tools which can be used to gain these insights. The base assumption is that each product can be divided into different product attributes or product characteristics like product features, design elements, or price.
Consumers compare products with these attributes to find and buy a product that suits them the best. These attributes vary from product to product and are an important factor that customers use to determine the value of those products. So, these variations are used by product managers to create unique selling propositions (USPs) and find a product market fit.
With conjoint analysis a product managers can:
- Better select valuable product features for implementation
- Assess the right pricing strategy for a product
- Compare your own product with the competitors’ products
- Optimize the marketing and positioning of the product
- Find the right target customer groups and market segments
Key elements of conjoint analysis
In conjoint analysis, a product is broken down into its attributes and characteristics. The product manager identifies the attributes that are of the greatest interest for the conjoint analysis and collects the characteristics of these attributes for his product and competing products.
The attributes can include product features, design elements, prices, and brand names.
As these attributes differ between products, these differences can be used in customer surveys to identify customer preferences and gain insights for product development.
The product manager defines these differences per attribute in a set of levels like:
- Attribute — Size
- Levels — Small, medium, and large
Each product is listed in product profiles and presented to potential customers in surveys with specific questions.
The type and style of how these surveys are built differ depending on the type of conjoint analysis you choose.
There are:
- Traditional conjoint analysis — Here Respondents rank or rate scenarios
- Choice-based conjoint analysis (CBC) — Respondents choose their most preferred scenario from a set of multiple-choice scenarios
- Discrete choice conjoint analysis (DCC) — Similar to CBC, respondents choose one preferred scenario from a limited set of options
Conjoint analysis example
Here is an example of a simple conjoint analysis comparing three different recruitment apps:
- App A
- App B
- App C
We will consider four attributes:
- User interface
- Job listings
- Resume builder
- Price
Each attribute has up to three levels:
- Average — 0
- Good — 1
- Excellent — 2
Below is a table visualizing the three profiles of the apps:
The respondents will be asked to rank these apps in order of their preference, from most preferred to least preferred. For example, their ranking might look like this:
- App B
- App C
- App A
Another respondent’s ranking might be:
- App C
- App A
- App B
After collecting rankings from multiple respondents, the data will be analyzed to determine the utility values for each attribute level and the overall preference for each app. The results will help identify which attributes are most influential in driving app preferences and which app is most preferred overall by respondents.
The analysis of the data is a mathematical process. Analyzing conjoint survey results is complicated and prone to measurement errors. Often participants don’t know exactly why they choose one thing over the other.
Survey results can induce substantial bias in any direction and by any amount; this bias must be corrected with mathematical processes. Econometric and statistical methods are used to estimate a utility function for each attribute and level of the attribute.
These utility functions indicate the perceived value of the attribute and show how consumer preferences are prone to change when the level of the attribute changes.
How to design and execute a conjoint analysis study
To design and execute a conjoint analysis study, you must be clear about the objective of the research. Depending on the objective and the complexity of the questions, the study needs to be designed in different ways and different conjoint analysis types can be chosen.
The desired outcome could offer insights such as:
- Identifying customer preferences
- Optimizing feature sets
- Understanding pricing sensitivity
After setting the objective the product manager must:
- Define attributes and levels of each attribute — It’s a best practice to not model too many attributes per profile. Keep it between 3 to 10 attributes and 3 to 5 levels per attribute
- Design a choice set of products to provide in a survey — To not overwhelm respondents in the survey, keep the sample set small
- Design a survey questionnaire matching the preferred conjoint analysis — Based on the format, it’s necessary to use proper tooling for this step. This is especially for dynamic surveys
- Execute the survey to collect data — Try to understand the respondents’ demographics and filter out respondents who do not suit your target group right at the beginning
- Analyze the data using a proper tool — Most tools use mathematical models and methods like hierarchical Bayes estimation
- Calculate the part-worth utility value — Use a tool to understand the preference values of each attribute level
- Interpret the results with consideration of the research objective — Find out which attributes determine the preference of a profile the most
- Act on the results — Define a feature set for your product and alter the pricing model accordingly
With the results of the conjoint analysis survey and the mathematical model in the background, you can even use the model to simulate how the preference will change for a certain product when attribute levels are changed.
More great articles from LogRocket:
- How to implement issue management to improve your product
- 8 ways to reduce cycle time and build a better product
- What is a PERT chart and how to make one
- Discover how to use behavioral analytics to create a great product experience
- Explore six tried and true product management frameworks you should know
- Advisory boards aren’t just for executives. Join LogRocket’s Content Advisory Board. You’ll help inform the type of content we create and get access to exclusive meetups, social accreditation, and swag.
Conjoint analysis case studies
Conjoint analysis is not bound to physical products. It can be used across all industries for physical products and services alike. You can also use it for different scenarios like identifying the right price or improving the product and service offering.
The following case studies illustrate how you could use conjoint analysis:
A fitness equipment manufacturer
A company producing rowing machines was using conjoint analysis surveys to evaluate which features are most important for younger consumers who want to stock up on their home gyms. They asked different questions about the features of a new rowing machine model and found out that younger buyers would like to have rowing machines with the following attributes:
- Easier folding probabilities
- Touchscreens to play videos and see health metrics
- On-demand virtual coaching classes
- Silencing technology to keep noise level down
- Affordable compared to other competitors
They found out that different groups prioritize different things first. One group preferred convenience and quiet use, another prioritized high-tech interactive features, and another mainly looked at the price.
Better mousetraps
To optimize their products and services, a big pest control company used conjoint analysis to gain insights into the demands of modern customers. They asked in a survey what an improved mousetrap should look like. They figured out that:
- The trap should have an alert system synchronized with the smartphone
- The trap should send out low-frequency beeping sounds to keep mice outdoors
- There should be a mercy model which catches the mice without harming them
- There should be a subscription-based carefree service model where someone comes to your house and maintains the traps
After developing some of these insights customer satisfaction increased.
Improved phone service
A phone answering service improved its offering by using a conjoint analysis based on the following attributes:
- On-demand, personal answering support
- Follow-up communication via text and social media messaging
- A pay-per-call model
- A company calendar management service
- An up-selling consulting service that helps customers to up-sell their products via phone
They identified easy ways to improve their services with little effort but with great value increases for their customers.
Advantages and disadvantages of conjoint analysis
Designing conjoint studies is complex. When too many product features and product profiles are chosen, respondents may often feel overwhelmed and tend to simplify the answers to questions.
The mathematical model that supports conjoint analysis is also very complex. The results and the way they’re calculated may not be easy to understand and interpret.
When conjoint analysis studies are poorly designed, they may overvalue product attributes which trigger emotional responses, and undervalue concrete features and important hard facts.
In the survey, the respondents are presented with all the attributes of a profile. In real life, the product positioning is harder and the consumers seldom have all the facts presented in this way. The conjoint analysis can therefore only be a reference and not directly put into practice.
On the other hand, conjoint analysis has numerous advantages. Above all, the fact that psychological mechanisms play a role in decision-making in conjoint analysis is an advantage. After all, emotions also play an important role in the real buying process.
In addition, conjoint analysis presents several attributes to the respondent in a combined manner, which corresponds better to reality than a survey in which individual attributes are queried.
In addition, conjoint analysis relates the various attributes to each other, which means that the most important factors for the user’s preferences can be identified.
Tools to design and execute conjoint analysis
There are plenty of tools out there that support product managers and market researchers with their conjoint analysis. The following are the most common:
- Sawtooth Software — Sawtooth Software is one of the most widely used and comprehensive tools for conjoint analysis. It offers various conjoint analysis techniques, including CBC, ACA, and MaxDiff. Sawtooth Software provides both standalone software packages (like Lighthouse Studio) for advanced users and online survey platforms (like Discover) for more straightforward studies
- SurveyMonkey — SurveyMonkey is another widely used online survey tool that supports conjoint analysis. While it may not have advanced conjoint analysis features like Sawtooth Software or Qualtrics, it can still be used for basic conjoint studies
- Conjoint.ly — Conjoint.ly is an online platform dedicated to conjoint analysis and related research methods. It offers automated conjoint analysis and simulation tools to analyze results and derive insights
Final thoughts
With the right preparation and a good selection of attributes and levels, conjoint analysis can give a product manager helpful insights into consumer needs. It can be used for pricing and competitive product analysis. At the same time, conjoint analysis can provide helpful insights into consumer behavior during the initial market research for a new product.
Due to the simplicity of the survey, there is no obstacle for the survey participants to take part in the conjoint study. Participants are only called upon to compare different profiles, which closely simulates a real purchase process. As a result, psychological mechanisms that play a role during the buying process are also included and flow into the conjoint analysis results.
Featured image source: IconScout
LogRocket generates product insights that lead to meaningful action
LogRocket identifies friction points in the user experience so you can make informed decisions about product and design changes that must happen to hit your goals.
With LogRocket, you can understand the scope of the issues affecting your product and prioritize the changes that need to be made. LogRocket simplifies workflows by allowing Engineering, Product, UX, and Design teams to work from the same data as you, eliminating any confusion about what needs to be done.
Get your teams on the same page — try LogRocket today.