When engaging in market research, information about the importance of various facets of your product or service can be hard to come by. Using choice-based conjoint (CBC) analysis, you can get a better grip on which specific elements really matter to your potential customers: like price, screen size, and warranty length.
These are the factors that customers really think about while making their buying decisions. Surveys that incorporate choice-based items are a real lifesaver for researchers who would otherwise have to rely on anecdotal evidence or best guesses about the most desirable characteristics of their offerings.
Overview of Choice-Based Conjoint Practices
Choice-based conjoint models rely on asking users to choose between a limited set of alternatives. These fictional products or services each have a number of stated characteristics, some of which might be the same and others different. For instance, as part of your market research, you could present your audience with the following choice:
Laptop 1
- 2.0 GHz processor
- 1 TB hard drive
- 6 hours battery life
- 15″ screen
- $1,200 price
Laptop 2
- 2.2 GHz processor
- 1 TB hard drive
- 4 hours battery life
- 15″ screen
- $1,200 price
The two laptops are almost identical, but Laptop 1 has a longer battery life while Laptop 2 has a faster processor. Survey-takers indicate which device they prefer, and then move on to the next pairing. By collecting the responses from many people and analyzing them, you’ll be able to see how important the trade-off between battery life and processor speed is.
The conjoint framework incorporates more than just differences between a couple of attributes. When creating your survey, you’ll choose the number of attributes to include as well as various levels of those attributes. You’ll also create numerous distinct profiles of products. These profiles represent specific options with different levels of each of the attributes. Modern tools allow you to automate the process of profile creation.
The Advantages of Choice-Based Conjoint for Product Research
Simply asking users to state which factors are important to them may do more to unearth consumers’ hidden biases than reveal valuable marketing info. Consider the question, “Is environmental friendliness important to you when buying electronics?” Most people would like to think of themselves as eco-friendly, and so you’ll probably get a lot of “yes” responses to that question. But, how much does eco-friendliness really influence a consumer’s choice? The answer may not be as clear-cut. By including environmental friendliness as just one attribute among many, the conjoint way of doing things helps to minimize this skewing of the data.
Another benefit of this type of survey is that it gleans data pertinent to actual considerations that consumers take into account when evaluating the characteristics of real-world products. You can take effective action upon finding out that most laptop buyers are willing to trade 0.2 GHz of processor speed for an extra 2 hours of battery life. If you instead asked participants to rate how important processor speed and battery life are on a scale from 1 to 10, you might well discover that processor speed rates an average of 7.6 while battery life clocks in at 6.2. Good luck trying to use this trivia to sell your machines.
In most cases, you’ll have more than two levels of each attribute. This means that you can see how much people value that attribute at higher and lower values. You may find that your customers really like having a 4-year warranty rather than no warranty at all, but they don’t care much for extending the warranty to 10 years. This knowledge will allow you to devote your resources to where they will make the biggest impact on sales and customer satisfaction.
But, why not have respondents rank all possible combinations at once?
Early versions of research involving the use of conjoint decision-making asked individuals to rank many options against each other.
While this produces nice graphs of the results, it’s not the way that people actually act in shops or at e-commerce sites. Nobody, when trying to buy dinner, sits there and ranks 50 meals from 1 to 50. They instead consider a few choices, perhaps even just two, and compare them against each other.
Choice-based conjoint mimics this natural mental process to give you more accurate results.
Choice-Based Conjoint Analysis and Price
Many of the elements of a product or service can really only be selected from among a few alternatives. Automobiles come in two-door and four-door models, plus the possibility of a hatchback as a third variant. Price, on the other hand, is a very continuous spectrum. This is why many enterprises segment their product lineups to cater to consumers at various price points.
Using the power of choice-based conjoint analysis, you can conduct accurate pricing research. By experimenting with higher and lower prices, you can see which features and characteristics people expect and will pay for at differing levels of cost.
Other types of conjoint analysis, such as adaptive conjoint analysis and full profile conjoint analysis, are widely known to be ill suited for the purposes of pricing research, so the choice-based conjoint model is your best bet and the industry favorite.
Sophisticated Number-Crunching
Due to the nature of choice-based conjoint analysis, it doesn’t lend itself to quick, back-of-the-envelope calculations.
To tease out the insights about customer utility preferences hidden in the survey responses, quite advanced statistical methods are required. They may include multinomial logistic regression, maximum likelihood estimation, Bayesian techniques and other advanced algorithms.
Fortunately, modern analysis software means that you won’t have to try to perform these mind-numbing computations by hand.
Designing a Good Choice-Based Conjoint Survey
There are a lot of pieces to play around with when it comes to surveys using choice-based conjoint items. This gives you plenty of flexibility to craft a study that suits your needs, but it also means you can easily make a mistake that will invalidate your results. Follow these tips for the best chances of creating a successful choice-based conjoint survey:
1. Limit Your Attributes
There are probably dozens of characteristics that differ among your offerings and those of competing businesses. It’s important to narrow down the list to only those that are relevant. While purchasers might have a slight preference for a computer monitor with four buttons rather than three, this small detail may play almost no role in their buying decisions.
Overloading your survey with extraneous elements will make it more difficult to identify those that are important. Another drawback lies in the fact that too many attributes will make your survey harder to read and may lead to fatigue among participants.
Try to select your attributes in such a way that they’re fully independent of each other. Otherwise, you might incorrectly interpret data that suggests that a certain attribute is strongly preferred when, in fact, it’s a related attribute that’s highly desired by consumers.
The best way to do this is to conduct surveys with a sample of your target audience to determine which attributes they care about. Doing this before you dive into your full study will save you time and money in the long run!
2. Choose Realistic Levels
Select the levels of your attributes in ways that make sense for the industry you operate in and the demographic you’re targeting.
Suppose you intend to debut a low-priced wristwatch in the near future. Well, despite the fact that Rolex offers some models for $5,000 or more apiece, you should not include this price as one of your levels. You could instead have $10, $20 and $50 as your levels when engaging in pricing research.
Setting inappropriate levels in your survey (in this case, high prices) might provide you with some fascinating and curious findings, but almost all of them will be useless from a marketing research perspective.
3. Spend the Time to Do It Right
Because of the large number of variables and considerations you have to deal with when designing a choice-based conjoint survey, it really pays to do your work thoroughly. You may even wish to implement an iterative process wherein you add numerous attributes and levels and then whittle them down with each step to a smaller, more manageable number until you’re satisfied with them.
Use any information you’ve received from previous market research to inform your ongoing survey design practices.
As we mentioned in step 1, you can create a few preliminary surveys to gauge the right levels and attributes to include in your “real” one. While it may seem like an additional investment of time and money, doing your homework before the big test always pays off.
Connect With Customer Preferences
Employing choice-based conjoint analysis in your marketing research may enable you to reveal truths about customer preferences that you had not suspected before.
Particularly when it comes to pricing research, conjoint analysis using a choice-based methodology likely has a crucial role to play for your organization.
Oh, and remember to do your homework diligently when selecting attributes and levels for solid, reliable results.
Interested in using Choice-Based Conjoint in your next project? Sign up for a free trial of Alchemer Team Edition.
