5 Real-World Examples of MaxDiff Analysis

Introduction to MaxDiff 

Key to many marketing research, economics, and UX studies is learning people’s preferences, including what’s important or not to them. MaxDiff (Maximum Difference Scaling), also known as best-worst scaling, is a research method that helps organizations measure preferences/attitudes and prioritize a list of items. Unlike traditional rating scales, where respondents often provide lazy answers such as saying that most everything is important, MaxDiff requires respondents to choose between options, leading to clearer and more accurate insights. 

Sawtooth Software, a global leader in survey research tools, offers easy-to-use MaxDiff solutions to help businesses, researchers, and institutions leverage this valuable methodology within their questionnaires. From prioritizing product features to understanding consumer choices, MaxDiff is a trusted method for making informed decisions.

What Is MaxDiff Analysis?

MaxDiff Analysis is a choice-based survey approach for estimating the relative preference (or importance) of a list of items. Respondents see sets of typically 4-6 options per choice set, and pick the one they like most and least within each set. 

The results lead to a prioritization on a numeric scale (often summing to 100) that better reflects respondents’ true preferences than measurement approaches involving rating scales. 

MaxDiff works better for cross-cultural research, because it avoids scale use bias. Scale use bias is when different people, often from different cultures, use the rating scale differently (e.g., people from India may use more 4s and 5s than people from Germany). MaxDiff leads to directly comparable preference scores across people and groups of people with different scale use tendencies than rating scale data.

By reducing bias and improving clarity, MaxDiff provides more valid and reliable insights than rating scales. Because it does a better job capturing information, you can obtain better results with smaller samples sizes than when using rating scales. MaxDiff is used widely across industries—from tech companies refining product features to tourism venues improving guest experiences. 

A recent review of 526 academic journal articles by Schuster et al. highlighted the wide use and acceptance of the MaxDiff technique over a variety of applications including economics and marketing.

Below are five real-world examples of how organizations use MaxDiff Analysis, as reported in Keith Chrzan and Bryan Orme’s book, “Applied MaxDiff: A Practitioner’s Guide to Best-Worst Scaling”, published by Sawtooth Software.

5 Examples of MaxDiff Analysis

Using MaxDiff to Prioritize Features at Google

At a recent Sawtooth research conference, Chris Chapman and Eric Bahna reported on a valuable way that Google used MaxDiff to solve real-world challenges. While MaxDiff is often used for product features and marketing messages, they applied it to prioritize feature requests.

Prioritizing Feature Requests

Feature requests come from customers, executives, and engineers. Product managers must choose which to develop, balancing importance with cost. This is similar to the “knapsack problem,” where multiple smaller features might offer more value than one large feature.

Understanding Differences

Google’s use of MaxDiff prevents decisions from being based solely on executive opinions. They have found that product managers and sales teams often prioritize requests differently. This data helped create strategic discussions and smarter backlog planning.

Considerations

Challenges included low survey participation and varied understanding of features. When Chapman and Bahna showed participants that answers were used to change the product roadmap, participation went up and participants expressed enthusiasm about the process. As a result, Google continues using MaxDiff to enhance decision-making across teams.

Delighting Customers at Riot Games Utilizing MaxDiff

Kegan Clark, Senior Researcher at Riot Games, shared how MaxDiff helps product managers understand player feedback and game features. Riot Games also applies MaxDiff in unique ways beyond traditional research.

MaxDiff improves comparisons across regions (e.g., USA vs. China) by removing cultural bias (e.g., scale use bias). When evaluating dozens of game updates, MaxDiff helps researchers at Riot identify with greater precision and discrimination the most frustrating issues for players.

In past surveys using Likert scales, responses showed little difference between options, making it hard to interpret results. By switching to MaxDiff, Riot Games found strong links between major frustrations and lower playtime, allowing them to prioritize the biggest issues affecting player retention.

MaxDiff Analysis at the Polynesian Cultural Center

The Polynesian Cultural Center (PCC) in Laie, Hawaii, has been a leading attraction for over 55 years. To improve visitor experiences and marketing, the PCC uses Sawtooth Software’s MaxDiff for precise insights.

How MaxDiff Helped:

• Visitor Segmentation – A study ranked 27 travel-related statements, identifying three visitor types with distinct booking behaviors.
• Feature Value Assessment – MaxDiff helped align ticket prices with guests' perceived value.
• Gift Bag Preferences – Results led to swapping DVDs for bakery coupons, doubling perceived value and increasing retail sales.
• Marketplace Expansion – MaxDiff and TURF analysis guided the selection of new food vendors, boosting marketplace growth.

The PCC continues using MaxDiff to improve visitor satisfaction and business success.

Understanding E-Cigarette Preferences and Attitudes with MaxDiff

Researchers at the University of North Carolina’s Department of Family Medicine conducted a MaxDiff study to identify the factors that most and least encourage e-cigarette use. They surveyed U.S. adults (18+) who had tried e-cigarettes, asking them to complete 19 MaxDiff tasks. They evaluated nine key factors, such as health risks, nicotine content, and price.

UNC researchers chose MaxDiff because it simplifies decision-making. Compared to other methods they considered, it reduced cognitive strain and ensured clearer results. The study showed that e-cigarette users consider multiple factors when making choices, offering precise insights into consumer behavior and strategies to improve overall health, including messaging.

MaxDiff at Progressive Insurance

Progressive Insurance, known for its data-driven approach, needed to evaluate a considerably large list of nearly 100 insurance benefits to guide marketing strategies. They used a sparse version of MaxDiff survey to handle this large list efficiently.

Sparse MaxDiff allowed Progressive to test all benefits without overwhelming respondents. (With sparse MaxDiff, each item is shown typically just one time to each respondent: for example, 19 sets of 4 items allows the questionnaire to cover each of 96 items for each respondent). The MaxDiff format simplified choices, reducing biases common in traditional rating surveys. This also made it easier to explain results to internal stakeholders, increasing trust in the methodology.

The study’s results prioritized benefits by importance. Using TURF analysis (an optimization algorithm that can be applied to MaxDiff data), Progressive also tested 1000s of potential benefit packages, helping them optimize marketing and product strategies.

Since then, Progressive has regularly used MaxDiff across different customer groups, proving its value for ongoing research and decision-making.

Conclusion

MaxDiff Analysis is a powerful tool for organizations looking to make data-driven decisions. Whether it’s prioritizing product features, understanding consumer preferences, giving insights into health-related issues, or refining marketing strategies, MaxDiff delivers insights in a more precise and unbiased way than traditional rating-scale surveys.

Sawtooth Software leads the way in MaxDiff research, offering easy-to-use survey tools that help organizations collect high-quality data. These real-world examples show how MaxDiff can drive success across various industries, making it an essential tool for better decision-making.