Conjoint analysis is a marketing research method that leverages statistical analyses and mathematical models to quantify survey respondents' preferences for product features, determine what attributes impact product demand, and can also predict how the market will respond to new products pre-launch.
These research methods use specialized surveys to ask respondents to pick between bundles, and the importance of specific product features is then analytically derived from the results. Once the survey results are collected, the organization can determine how valuable each product feature is and if they should implement them before launch.
For example, a TV manufacturer might conduct a survey where they give the respondents bundles with different features to pick from. Each bundle might include screen type, size, brand, and price. Respondents will choose which bundles are more appealing to them, and the TV manufacturer can determine how much each respondent values certain features.
Conjoint analysis is becoming increasingly popular among market researchers and other organizations because it provides valuable insights that are difficult to obtain elsewhere.
Conjoint analysis helps organizations optimize features and pricing by giving them the means to quantify and study consumer preferences. Using conjoint analysis, organizations can measure the importance of different factors by having customers choose between realistic product packages. Organizations can deduce the importance of various factors, such as pricing and product features, by modeling utilities after the data derived from conjoint exercises. When using conjoint analysis in tandem with market segmentation, you can easily narrow down which customers prefer what features or services.
If you ask consumers directly what is important to them, the answer is usually everything.
For example, most respondents will say that price is very important when buying a smartphone. However, high-price brands like Apple have a very strong market share. Factually, price is less important than product features like the brand, the features, etc. Optimizing features and pricing requires more accurate quantification of consumer preferences. Respondents are not asked to state the importance of different factors. Instead, they are asked to pick between realistic options (products with features and prices). The importance is derived from the choices they made. This is why conjoint analysis is often used over other methods.
For market research agencies, purchasing survey software is an important decision. For example, many agencies evaluate survey software on three dimensions: functionality, price per complete, and quality of service. Conducting a conjoint analysis across these three dimensions provides insight into tradeoffs between the three.
At IntelliSurvey, we pride ourselves in our ability to immerse respondents in the conjoint exercise. Our video walkthrough of a survey conjoint with an agency’s lead generation activities is a great example
Effective conjoint analysis models feature natural relationships between respondents and product attributes. Conjoint exercises should appear “au naturel” to customers to yield the most accurate data possible. One way to accomplish this is to tailor conjoint models specifically for critical customer segments; various segments will weigh attributes differently, and taking a one-size-fits-all approach may invalidate the data.
The best conjoint analysis models incorporate natural relationships between respondents and product attributes. Effective conjoint analysis models feature surveys that closely resemble the final decision point; for example, if you conduct a conjoint analysis for phone plans, you might want to replicate a large phone carrier’s bundle price range.
If imperfectly contextualized, the presentation might not reveal natural attribute preferences. Over time, attention to detail on the look and feel has decreased for practical reasons; organizations run more conjoints, and directional results are often enough. However, the best conjoint results leverage UI elements close to the final product to obtain more detailed results.
Where applicable, organizations should tailor conjoint exercises for critical customer segments. For example, an airline may create different conjoint models for frequent and infrequent flyers. Regular airline customers may receive a free, first-class upgrade, so measuring first-class against economy will have to incorporate more factors than just pricing preferences.
For many organizations, it can be challenging to analyze these complexities and adapt their exercises to provide options for critical customers, but it is crucial for gaining valuable insights across all customer segments.
Roughly half of the online survey traffic is mobile, meaning voluminous choice options will not properly fit on a smartphone or similar device’s single screen. In these exercises, scrolling can become too intense for respondents to focus on and cause them to disengage.
It is possible to limit conjoint exercises to desktop users, but this creates a selection bias in survey responses. In many cases, acquiring data from younger respondents becomes challenging. Conjoint presentations have been streamlined, often with intro pages to give details to keep the task manageable on a mobile device.
It is critical for organizations to review and test conjoint exercises on an online device before launching to ensure proper data quality.
There are four different types of conjoint analysis methods that organizations employ during market research: choice-based conjoint (CBC) analysis, adaptive choice-based (ACB) conjoint analysis, adaptive conjoint analysis (ACA), and menu-based conjoint (MBC) analysis.
Choice-based conjoint (CBC) analysis is one of the most common forms of this research method. CBC records how a respondent values different combinations of features within a product or service.
This conjoint analysis method asks survey respondents to review a set of product concepts with potential features and allows them to select their favorite combination. Using these results, researchers can predict the market share for different scenarios, depending on the product they roll out.
Adaptive choice-based (ACB) conjoint analysis models are newer and more advanced than other conjoint models. Adaptive models are most effective when the attribute list grows because it delves deeper into the respondents' preferences. In return, this conjoint analysis method provides a thoroughly engaging experience for the respondent. ACB conjoint analysis models are more detailed and require more time to conduct than a CBC study.
Adaptive conjoint analysis (ACA) is a conjoint analysis research method from the 1980s that provides each respondent with a personalized experience. This conjoint analysis method gives each respondent a different survey experience based on their answers to previous questions. Although many organizations do not use ACA today, it is helpful where organizations must evaluate numerous product features or attributes simultaneously to hasten the process.
Menu-based conjoint (MBC) analysis models are advanced tools that assist in analyzing menu choice experiments with multiple checks. While this conjoint analysis model provides organizations with better opportunities for modeling complex consumer preferences, it requires a high level of expertise. Many organizations will defer this to experienced professionals with the necessary platforms to make this conjoint model scalable and viable.
Conjoint analysis methods help organizations conduct market simulations, predict demand, and estimate price sensitivities for potential products. Many researchers also leverage data from conjoint analyses to organize respondents based on what product attributes and features they prefer. In short, conjoint analysis is used to help companies determine which features users value the most and assists them primarily in the three following areas: pricing, marketing efforts, and research and development.
Conjoint analysis helps businesses determine pricing for different products and services. Using conjoint analysis research methods, organizations can compare various product features to determine how consumers value each. For example, attributes that users appreciate more in service might wind up in a more expensive subscription package, while lesser-valued features might become part of a trial version.
Conjoint analysis methods help organizations create marketing strategies by informing them of which factors are most valuable to various target audiences. Organizations can leverage respondents’ answers about which features are highly favored to determine where they should allocate their marketing resources. Conjoint analysis can also help marketers segment and target different audiences based on what features different demographics value the most.
Insights obtained from conjoint analyses can assist an organization in researching and developing new products and services. For example, if a competitor's product offers more features, they can deploy a survey to their users with sample packages that include those features. If users favor one sample over the others, the organization can move forward with implementing those features.
Modeling is a critical part of running a conjoint project and has evolved over the years. Conjoint experts back in the day were much more attentive to the design and presentation of the choice scenes themselves. We regularly heard concerns about whether respondents would be able to interact with the scenarios in an intuitive way. If imperfectly contextualized, the presentation might not reveal natural attribute preferences. Thus, IntelliSurvey often undertook great efforts to model conjoints that would appear “au naturel” for respondents - in market freezer doors, new car stickers, six-packs, and more.
As conjoint has gotten more "packaged", one can see much more "plug and chug" thinking about the exercises. Presumably, for many exercises, this is just fine. When the “au naturel” target interaction involves respondents picking one plan from three presented in three columns. Such exercises are quite straightforward for data collectors to deploy, sometimes requiring just a few hours.
We're always jazzed to see more complex models, and presentation schematics. We're excited to work with friends old and new from the high-end modeling community, and introduce interested parties to new modeling methods. The world of bespoke still exists, and may be better than the cookbook at reliably modeling the utilities and relationships critical to your business success.
IntelliSurvey has a variety of products to help companies conduct surveys and multi-market studies. IntelliSurvey has been deploying conjoint analyses for more than 20 years and excels at creating models that are intuitive and organic to survey respondents. We help organizations optimize their features and pricing based on customer preferences, contact us for more information.