Many businesses struggle to understand what truly drives purchasing decisions. Traditional surveys can be misleading, as customers may not accurately report their preferences or may be unaware of the factors influencing their choices. This leaves companies guessing about which product features to prioritize, how to price effectively, and how to allocate resources for maximum impact. Insert - conjoint analysis.
Conjoint analysis offers a powerful solution to these challenges. It presents customers with realistic trade-offs between product attributes. It reveals the hidden factors behind purchase decisions and provides businesses with quantifiable data on customer preferences. This allows for more accurate product design, pricing strategies, and marketing efforts.
Top brands rely on this technique to perfect their offerings. It guides them in setting the right price, choosing the best features, and outperforming competitors.
The outcome? Products that sell fast and keep customers coming back.
In this article, I'll share some examples that highlight the powerful insights and strategic guidance that can be derived from this methodology, stepping through:
● What is conjoint analysis?
● Questions conjoint analysis answers
● Conjoint analysis example: daily tradeoffs consumers face
● Overview of how conjoint analysis works
● Ways conjoint analysis benefits businesses
● Tips for designing a conjoint-based survey
What is conjoint analysis?
Conjoint analysis is a technique that originated in mathematical psychology (Luce and Tukey 1964) and was first applied to marketing problems by University of Pennsylvania Wharton School Professor Paul E. Green (1971). This technique helps businesses measure the tradeoffs consumers make when choosing between alternatives.
Usually conducted within an online survey, a conjoint experiment breaks down products and services into their key attributes, like price, features, quality, brand, etc. Survey takers are then presented with multiple alternatives of the product or service that combine different levels of these attributes and, in a choice-based conjoint, are asked to choose their preferred option.
Researchers can determine the importance of each attribute, assign a value to them if needed, and determine how consumers make tradeoffs between them by analyzing people's choices.
Conjoint analysis is particularly helpful in marketing, product management/development, and operations research.
Questions conjoint analysis answers
Conjoint analysis provides uniquely actionable insights by quantifying customers' tradeoffs between product attributes. By forcing survey respondents to choose between product profiles, we can uncover what they truly value. This enables businesses to answer critical questions like:
How much more will customers pay for additional product features or higher quality?
Conjoint quantifies willingness to pay for additional features, quality, or benefits. You can determine how much more revenue premium offerings will generate or how much market share would be lost if prices increased.
Which attributes are deal breakers versus nice-to-haves?
Not all product attributes are deal breakers. Conjoint exposes which features customers will compromise on versus those they require. This avoids over-engineering products.
How price-sensitive are different customer segments?
Willingness to pay often varies across customer segments. Conjoint can identify price-sensitive segments based on the tradeoffs they make. This will help companies decide whether they need to have a budget offering, a premium offering, or both.
What is the revenue-maximizing combination of price, features, and quality?
A product with all the best features at the lowest price will certainly be preferred. However, this is typically not optimal from a revenue standpoint. With data from a conjoint model, we can use algorithms to search all possible product configurations. This helps find the set of features that attract enough people at the highest price they're willing to pay, maximizing the revenue generated for the business.
If cost information is provided, the recommendation could become a profit-maximizing combination. Knowing the configuration that will drive the most revenue and/or profit helps companies avoid underpricing or overpricing. Underpricing leaves money on the table. Overpricing can potentially lose sales.
Varying product configurations with conjoint identifies the price, feature set, quality level, and other attributes that maximize appeal. We can learn what people truly value because we force them to make trade-offs. The insights from conjoint analysis are unmatched by traditional survey techniques.
Conjoint analysis example: daily tradeoffs consumers face
When booking a flight, consumers compare departure times, number of stops, airline loyalty programs, and ticket prices. While the cheapest nonstop on a preferred airline is ideal, it is rare that a person finds the perfect combination and, oftentimes, needs to compromise based on what's most important for that particular trip.
Think back to the last time you bought a new smartphone. There are so many options to choose from. We must consider different brands, screen sizes, camera quality, battery life, storage space, and price. Since we likely can't maximize every feature within our budget, we select the phone with the combination of attributes that we value most.
Conjoint use cases
Consumers face tough choices daily. They weigh features, quality, and price when buying products and services. Business owners grapple with similar tradeoffs in design, pricing, and marketing. Traditional market research often misses the complex reasons behind these choices. Conjoint analysis fills this gap, providing a proven way to examine these crucial tradeoffs.
Let’s imagine you run an airline company. In order to be profitable, you'll have to balance ticket prices, flight schedules, aircraft sizes, amenities, and payouts of the loyalty programs. Having too few flight options or too many overpriced flight options could result in lost customers. But too many underpriced seats will hurt your profitability.
Or perhaps you work at a smartphone manufacturer. You can't change your brand, but you can build multiple SKUs, with different features and sizes and specs and material to try and attract more buyers. But too many SKUs with too few buyers will result in high manufacturing costs which may hurt your bottom line.
Manage a quick-service restaurant? You may want to understand the most desirable menu items, portion sizes, and pricing.
Having a lot of menu options requires more ingredients, kitchen space, and staff training to ensure that each dish is prepared consistently and to the restaurant's standards. Not to mention, a broad menu can lead to higher levels of food waste. But a limited menu may not cater to the preferences of your potential customers, causing them to dine elsewhere.
These examples just skim the surface! Businesses should consider conjoint analysis any time they want to inform their strategic decisions by consumer preferences.
Overview of how conjoint analysis works
First, identify the key attributes of your product or service that drive customer choice. These could be price, features, quality, or brand. Then, define levels for each of the attributes. For example, the price attribute may have three levels - $49, $79 and $99. The brand attribute may have four levels - AeroGlow Tech, StellarBlend, LuxoPeak, and ZeniFlora.
Next, you will need to generate an experimental design that combines the levels into different profiles. Each profile represents a product variant. In an online survey, you'll typically show respondents 3 or 4 profiles at a time. You'll ask them to pick their preferred option.
You'll repeat this process over a series of screens, typically 10-12. After collecting the data, you'll build a model that analyzes people's choices. This creates a utility value for every level of every attribute.
Conjoint analysis assumes the total utility of a product is the sum of its parts. You can use market simulations to determine which features your product should include. You can also find the optimal pricing for that product.
The power of conjoint analysis lies in its ability to quantify consumer trade-offs. Respondents share preferences they can't articulate in traditional questioning. They do this by being forced to choose between product profiles.
Types of conjoint analysis
There are several types of conjoint analysis. Each type has its own strengths and uses. Here are some common ones:
Choice-Based Conjoint (CBC) is the most popular. It shows people a set of product options and asks them to choose one. This mimics real-world shopping decisions.
Best-Worst Conjoint asks people to pick the best and worst options from a set. This method provides more information about preferences but requires more effort from survey takers.
Adaptive Choice-Based Conjoint (ACBC) alters the questions based on earlier answers. This can give more precise results, but it's more complex to set up and analyze.
Menu-Based Conjoint (MBC) lets people choose from a menu of features. This works well for product or service offerings where people may want to buy more than one option at a time or bundle different products or features together.
Each type of conjoint has its pros and cons. The best choice depends on your research goals and the type of product you're testing.
Outputs and visualizations
Conjoint analysis produces several types of outputs. These help businesses understand and act on the results.
Traditional conjoint outputs include part-worth utilities and importance scores.
Part-worth utilities reveal the relative preference of each level within an attribute. For a computer, you might find that a 15-inch screen has a higher utility than a 13-inch screen or that a Dell has a higher utility than a Lenovo.
Importance scores show the relative influence a feature might have on the overall decision. For example, you might find that price accounts for 40% of the decision, while brand name accounts for 20%.
However, at Numerious, we believe part-worth utilities and importance scores don’t tell the full story. They are typically shown as the average across all respondents, and we all know that “lies are in the averages”. In addition, they are directly impacted by the attributes and levels you test.
If you have a really terrible level within one attribute, it might make that attribute have a really high importance score - when in reality it’s just that no one wanted a product with that one level. Not to mention, utilities and importance scores are not calculated within a competitive context.
The best practice would be to use the market simulator tool.
Market simulators let you test different scenarios and see how changing features or prices might affect interest in your product or service. These preference shares show the predicted market share for different product configurations based on how you’ve defined the market.
You can then generate price sensitivity curves to display how demand changes as price changes.
Output from the market simulator is often presented as graphs or charts. Visual presentations make it easier to understand and share the results. And, since the simulator is providing you shares of preference, you can compare shares directly where a 20% share is 2x a 10% share. Stakeholders can easily consume and relate to this data (versus utility scores).
Ways conjoint analysis benefits businesses
Conjoint analysis empowers smarter decisions across industries by quantifying consumer trade-offs.
Here are some examples of high-impact business applications:
● Smartphone makers can identify the optimal combination of features, specs, and price points across their model lineup. This maximizes appeal across customer segments while minimizing cannibalization.
● Cloud software companies can quantify their customers' willingness to pay for different tiers of services. They can then set pricing and feature bundles to maximize revenue.
● Auto manufacturers can determine the relative impact of MPG, safety ratings, brand, and options on vehicle choice. They can then focus design resources on features that matter most to consumers.
● Consumer packaged goods companies can figure out which new product attributes will command higher prices or drive trial. These attributes might include organic, gluten-free, and fair trade.
The applications of conjoint analysis in business decisions are endless. Conjoint provides a robust framework to enhance product, pricing, and positioning decisions in any industry. It produces data-driven answers to questions that impact the top and bottom lines.
Tips for designing a conjoint-based survey
While conjoint analysis is a powerful technique, it does require careful design and execution. When designing your first conjoint experiment, consider limiting your attributes and levels to the main drivers of choice. Typically, between 5 and 9 attributes, each with 2 to 7 levels, are ideal.
Ask enough conjoint questions to enable robust analysis but not so many that respondents experience survey fatigue. Typically 10-14 questions is ideal, but you can always increase your sample size in exchange for increasing the number of questions.
Accommodate the entire range of prices. Will your product or your competitors ever go on sale? Be sure to include the absolute minimum, and maximum, either you or your competitor will be priced at. One should not plan to extrapolate consumer preferences outside the price range tested.
Use clear language and define any technical terms, but don't "seed the witness" by playing up the new feature. Consider a hybrid approach by adding a quick qualitative phase to test your descriptions and the respondent experience.
Quality data begets quality models
By simulating the tradeoff decisions people face in the real world, conjoint analysis provides unique and actionable data on consumer preferences otherwise hidden from view.
While designing and fielding a robust conjoint study requires expertise, the payoff for manufacturers, marketers, and product developers is immense. Conjoint transforms product and pricing decisions from a guessing game to a data-driven strategic capability.
Comments