Product management is all about trade-offs. Whether the objective is increased market share, profit margin or revenue, every product manager makes trade-offs—quality vs. cost, time to market vs. breadth of features, richness of the offering vs. ease of use, etc.
So, how do you know what the market wants? What market segments exist? What those segments prefer? What will they pay? In short, how do you know what trade-offs to make? The answer is to get the market to make the trade-offs for you. Not the entire market, of course, just a representative sample of the market.
By using conjoint analysis, you, as a product manager, can do just that: understand the trade-offs you should make by understanding the trade-offs your market will make. Then, apply your increased market insight to your revenue, profit or share objective.
Is conjoint analysis right for me?
Conjoint analysis has been successfully applied in many industries, such as Air Travel, Smart Phones, Computers, Financial Services, Health Care, Real Estate, and Electronics. If your job includes configuring a defined set of features for a product or service and the consumer’s purchase decision will be “rational,” conjoint analysis can help. If, on the other hand, your consumer’s purchase decision will be “impulse” or “image,” conjoint is not the right tool for you. If you’re a technology product manager, conjoint analysis is right up your alley.
Because conjoint analysis helps you understand your market’s preferences, you can apply it to a variety of difficult aspects of the job, including product development, competitive positioning, pricing, product line analysis, segmentation and resource allocation. “How should we price our new product to maximize adoption?” “What features should we include in our next release to take market share from our competition?” “If we expand our product line, will overall revenue grow, or will we suffer too much cannibalization?” “For which value-added features is the market willing to pay?”
For example, a technology company was feeling pressure from a lower cost alternative and debated lowering its own prices. Then, the results of a conjoint analysis showed the market valued their products differently from the competitors. They chose not to lower prices, but to slightly reconfigure their offering. As a result, the business grew and realized substantial profits that they otherwise would have never seen. Not every situation is as dramatic as that, of course, but a conjoint analysis done right is impactful.
Conjoint analysis is a set of market research techniques that measures the value the market places on each feature of your product and predicts the value of any combination of features. Conjoint analysis is, at its essence, all about features and trade-offs.
What exactly is conjoint analysis?
Conjoint analysis is a set of market research techniques that measures the value the market places on each feature of your product and predicts the value of any combination of features. Conjoint analysis is, at its essence, all about features and trade-offs. With conjoint analysis, you:Ask questions that force respondents to make trade-offs among features
- Determine the value they place on each feature based on the trade-offs they make
- Simulate how the market reacts to various feature trade-offs you are considering
To demonstrate conjoint analysis in action, let’s consider cell phone plans. These plans have various feature types, which in the language of conjoint analysis are called attributes. Let’s focus on Brand, Price, Minutes, Rollover Options, and Call Options. In reality, plans can be more complicated and conjoint analysis can keep up with the complexities, but let’s keep the example simple. Each of the attributes listed above has different levels. The levels of the Brand attribute might be AT&T, T-Mobile, Verizon, etc., but here we will refer to possible Brands as Brand A, Brand B, etc.
Brand A, Brand B, Brand C, Brand D
$60/month, $75/month, $100/month
800; 1,000; 1,400; 2,000
No rollover of unused minutes
No free calling based on contacts
Attributes must be something you can categorize, but they don’t have to be numeric. Note that the attributes include brand, price, and various product features. Through conjoint analysis, you gain insights into the value of your brand and the value of product features, and determine price sensitivity.
Survey the market
Conjoint analysis survey questions could take a variety of forms, depending on your study objective, but the most common type of question would be:
Which of the following cell phone plans do you prefer?
|Brand A||Brand B||Brand C|
|1,400 minutes||1,000 minutes||800 minutes|
|Unused minutes rollover for 1 month||No rollover of unused minutes||Unused minutes rollover for 1 year|
|No free calling based on contacts||Free calling to top 5 contacts||Free calling to top 10 contacts|
|Costs $100/month||Costs $75/month||Costs $60/month|
Derive values for each of the levels
From responses to these questions, conjoint analysis uncovers the underlying value for each level, depending on how often a level was included in the product selected. The relative value of the levels is what is relevant, in other words, how the value of one level compares to the value of another.
For example, the values for the levels of the Call Options attribute and the Rollover
Options attribute for one respondent might be:
|Call Options||Value||Rollover Options||Value|
|Free calling to top 10 contacts||50||Unused minutes rollover for 1 year||100|
|Free calling to top 5 contacts||20||Unused minutes rollover for 1 month||30|
|No free calling based on contacts||0||No rollover of unused minutes||0|
You can see in this example, given the levels tested (which is an important caveat), the Rollover Options attribute (with values ranging from 0 to 100) was more important to the respondent than the Call Options attribute (with values ranging from 0 to 50). These values can be calculated for individuals as well as for the overall market, which means you can use conjoint analysis to segment your market based on respondent characteristics, needs and preferences. Each of the level values is called a part-worth, because they represent the worth of any given part of the product.
Predict preference for various products
Once you see the part-worths, you understand what trade-offs to make so a product will be more desirable to the market. This predictive capability is where the real power of conjoint analysis is evident. For example, given a set of part-worths, you might have the following scenario:
|Brand A||30||Brand C||0|
|1,000 minutes||55||1,000 minutes||55|
|Unused minutes rollover for 1 month||30||No rollover of unused minutes||0|
|Free calling to top 5 contacts||20||Free calling to top 5 contacts||20|
The total value of the Brand A product is 30 more than the Brand C product. This consumer would be more likely to select the Brand A product. But, if the Brand C call option was changed from “Free calling to top 5 contacts” (part-worth of 20) to “Free calling to top 10 contacts” (part-worth of 50), the overall value of each product would be the same and the consumer would be equally likely to select either product. The overall value of a product is referred to as its total utility.
Simulate competitive markets
Now that each attribute level has an associated part-worth, we can create any number of competitive scenarios by mixing and matching the levels and increasing or decreasing the number of products. The result of any conjoint analysis study is a simulation model that allows you to simulate, for example, what share of the market will prefer your product versus your competitors’ products. For example, you might see results like this:
|Brand A||Brand B||Brand C|
|1,000 minutes||1,400 minutes||1,000 minutes|
|Unused minutes rollover for 1 year||Unused minutes rollover for 1 month||Unused minutes rollover for 1 year|
|Free calling to top 5 contacts||Free calling to top 10 contacts||No free calling based on contacts|
|40% share||35% share||25% share|
These shares, totaling 100%, are called “shares of preference,” because they refer to the share of the market that prefers each product, if everything else were equal. They are not market shares, because they don’t take into account a variety of other factors, such as sales and marketing efforts, distribution channels, product lifecycle phase, etc.
Simulating shares of preference is powerful. And, there’s no limit to the simulations you can run. So, for example, if your competitor changes its product, you can run simulations to help determine your response. If you are contemplating adding a new product, you can predict whether that will be beneficial and from which product in the existing market your new product will grab the most share. These are simple but potent examples of the many different ways that conjoint analysis may be used. To get a complete picture of the competitive landscape, include all competitors, and to ensure the predictive capability of the approach over time, think carefully at the start about the attributes and levels you will include in the study.
That said, because it allows endless scenarios to be tested in a competitive landscape, share of preference allows for powerful “what if” analysis. It ultimately provides the insights you need to make the trade-offs you are faced with every day as a product manager. The insights gained regarding how you might change your position in the market, respond to competitive threats, grow revenue, penetrate specific segments, etc. can have a dramatic impact on the success of your product.
Analyze purchase likelihood
Even if your product is so new it has no competition and will create its own market, conjoint analysis provides powerful insights. In addition to market simulations and shares of preference, conjoint analysis also analyzes your product’s purchase likelihood. Purchase likelihood analysis uses the total utility of a product to determine a percentage indicating the relative likelihood that the product will be purchased, given various combinations of features and pricing.
Because purchase likelihood is single product focused and does not take into account the competition, it is particularly helpful when launching a product that is completely new to the marketplace. Purchase likelihood is often appropriate for micro-level product design as well, when major product decisions are already made, and the focus is on getting the details right.
What’s the best way to move forward?
There is software to help you design, conduct and analyze a conjoint analysis study yourself. But there are also nuances and important decisions to make in each step of the process. For example, there are different conjoint methodologies, each with its own approach to data collection. The one that is appropriate for you depends on the objectives of your study. Unless you’re going to personally do a conjoint study at least a few times a year, it’s likely that you will want to engage someone with experience in the field to help you navigate these nuances. Although surprising to many, the person you engage for your study need not be an expert in your field. You are that. They need only be expert in applying conjoint analysis to real business issues.
Just remember, the next time you’re making trade-offs as a product manager, use conjoint analysis to get your market to make the trade-offs for you.