A Better Way to Measure Customers’ Willingness to Pay

Set the price too high, and you may fail to get any takers. Set your price too low, and you might achieve plenty of demand, but you’re leaving money on the table. Neither outcome is ideal, and marketers have long been tasked with determining consumers’ “willingness to pay” (WTP)—the maximum price a customer will spend to obtain a product or service.

“At its core, this is a classic problem,” says Derek Rucker, a professor of marketing at Kellogg. “In fact, it’s among the oldest and most important problems we face as marketers.”

But when Rucker and fellow Kellogg marketing professor Eric Anderson, along with Kellogg alumna Sharlene He, now a faculty member at Concordia University, dug into the literature, they found that the prevailing consensus on how WTP should be conceived of was in fact vague and missing critical components.

Moreover, most of the existing methods of measuring WTP had serious downsides. “Current methods don’t consider context and competition the way they should,” says Rucker. “That means it is possible marketers could be making business decisions off inaccurate estimates of willingness to pay.”

So Rucker, Anderson, and He sought to find a better approach to measure WTP. In a new paper appearing in the Journal of Marketing, they do just this, introducing what they term the Comparative Method of Valuation, or CMV, and validating it across thirteen studies.

A better tool

Over the decades, a number of different approaches for measuring WTP have been developed. But they all have limitations in their ability to capture willingness to pay based on phenomena that marketing scholars know to be true.

Previous research has often found differences in people’s WTP across different contexts—for example, people are willing to pay more for a beer at a hotel bar compared with a beachside vendor. Yet many studies have measured WTP by simply asking people what amount of money they would spend on a product, without providing any details about the situation in which it is being purchased or the available alternatives. “In this simple example, you can see the problem,” says Eric Anderson, “You might pay upwards of $10 for your favorite beer at a fancy hotel bar, but only $5 for the same beer from a beachside vendor. That difference in WTP may be completely rational and driven by what alternatives are available. Just asking consumers what they are willing to pay misses a critical aspect of the metric, which is the comparisons consumers are considering.”

Most approaches to measuring WTP present customers with a product or service in isolation, as though their option is to either purchase it or go home empty-handed. While there certainly are some shopping situations where consumers will opt to leave with nothing, much of the time the question is “this or that” rather than “this or nothing.”

Rucker learned that lesson firsthand when his teenage daughter needed a dress for her school’s homecoming dance. Not getting a dress was never an option—it was always a question of which dress.

Even approaches that do consider alternatives don’t always take into account the fact that the most relevant alternative will differ from one situation to the next. Is the best comparative option that red dress for $75? That blue one for $45? The one you already have in your closet?

To illustrate the importance of accounting for the relevant comparison, consider how it might differ across customer segments. A consumer who is considering subscribing to Disney Plus will likely have a different comparison in mind—and therefore a different WTP—if they are currently subscribed to Netflix versus if they currently have cable. The right comparison is fundamentally tied to a company’s intended target segment and positioning strategy.

Comparisons matter

Enter the Comparative Method of Valuation.

The CMV builds on another popular approach, known as the Becker-DeGroot-Marschak (BDM) method. With BDM, consumers are asked to provide a measure of their WTP for a target product—say, a KitKat bar—and then see a random number. If that number is lower than or equal to their WTP, they must buy the KitKat bar. If it is higher, they get nothing.

This procedure has the advantage of keeping participants honest about how much that Kit Kat is actually worth to them—but it falls short for all the reasons described previously.

So the CMV adopts this random-lottery aspect of BDM, but then expands upon it.

“What’s fundamentally different about our procedure is that rather than simply asking whether you would buy or not buy something, we first want an understanding of who the competition is,” says He. “We want to ask, if you were not to buy this product, what would you buy instead? That’s going to be the relevant comparison shaping someone’s WTP. We don’t assume that the answer is they would buy nothing at all.”

In practice, this means introducing a few steps before the lottery procedure kicks in. First, participants choose from among several comparative options. If the target product is a KitKat, for instance, participants will choose whether to compare it against M&Ms at $1.10, Reese’s at $1.05, and Snickers at $1.15. This setup guarantees the target product is being compared against a viable and realistic alternative for that consumer.

Next, participants give an initial estimate for WTP by answering the question: “What price for the KitKat would make you equally happy to buy the KitKat for this price or buy the M&Ms for $1.10?”

Then comes the lottery procedure, where participants put their money where their mouth is. The distribution of all the resulting WTP figures, across a wide range of contexts, can thus be used to inform pricing.

One advantage of CMV is that it allows marketers to see which alternatives consumers are considering. The methodology is also flexible and can be customized to different situations. For example, if a marketer wants to understand the WTP for KitKats in the constrained context of a vending machine, they can test it against only other candy bars; if they want to understand it in the context of a grocery store overflowing with snack choices, they can test it against other candy bars, as well as (say) apples and crackers.

Putting Comparative Method of Valuation to the test

The researchers conducted over a dozen empirical studies to test the CMV as a tool and explore its managerial applications.

The empirical studies took them several years to run and involved almost 3,000 participants—in person and via online platforms such as Amazon Mechanical Turk. “We learned different things in each study,” He says. “Each tackles the problem in a different way.”

For example, in one study, the researchers tested WTP for a premium version of Kraft Lunchables that includes a drink. In this case, the company was trying to convert customers of existing Lunchables to consumers of the premium product. To that end, the researchers tested the premium version against three flavors of regular, no-drink Lunchables, all priced at $1.36 (the price at Walmart at the time of the study). They also included a “no-buy” option, designed to identify consumers who were probably unlikely to purchase any Lunchable product and therefore not realistic customers. As Anderson notes, “One neat thing is we ran studies that reflect actual situations and problems marketers are trying to solve!”

Among consumers who selected one of the basic varieties, average WTP for the premium version was $2.33, while among the no-buy group, WTP was only $1.35—lower even than the price of regular Lunchables. Thanks to CMV’s ability to identify these different segments (that is, “buy” and “no-buy” groups), the researchers gained important insight into why some consumers had much lower WTP than others. In other words, having visibility into the comparative option a consumer is considering makes it easier to identify the true WTP—in this case, $2.33—of realistic customers.

CMV also proved to be a useful method for exploring situational factors like gift-giving, or for assessing how changing a product warranty might affect people’s WTP for different luxury refrigerator brands.

Notably, the researchers also ran an experiment where they tested CMV against its predecessor, the BDM method. When the WTPs produced by these two methods were compared in a real-world test of purchasing behavior, CMV outperformed its predecessor. “To be clear,” He says, “it is not that BDM isn’t useful, but we have improved upon that method to offer something with greater precision.”

The researchers also demonstrated when CMV outperforms conjoint, another popular method of estimating WTP. While conjoint is useful as an efficient way to explore WTP across many different contexts, CMV is superior when there are clear comparative options.

A useful tool

CMV contributes to the growing academic literature on behavioral economics, illuminating how different situational goals or comparative options can directly or indirectly affect customers’ willingness to pay.

But it can also be immediately adopted by marketers. “As a marketer, you need to not only understand what your brand offers—you also need to understand who you’re competing against,” Anderson says. “We can’t become so brand-centric that we just focus on ourselves. We have to know what alternatives consumers are bringing to mind.”

In future work, the researchers plan to expand their focus by investigating how a brand’s own advertising and that of competitors affect customers’ willingness to pay for a product—a direction that should offer even more practical insights for marketers. All three authors expect they will continue to share new findings on WTP with both MBA students and executives, who in turn can continue to test theories in real-world contexts.

“By no means have we solved the problem where we can tell someone with 100 percent accuracy the right price for a product,” Anderson says. “But we’ve taken a step in the right direction by clarifying the definition of WTP and giving people a methodology that directly takes into account comparison and context to better approximate real-world purchase situations.”