Getting There

The discovery of the Wallet Allocation Rule was made by a team of leading academics and industry experts with the goal of bringing the best of the science and practice together. Most important, the team wasn't just capable of doing a rigorous analysis. Its members were willing to go wherever the results led them—even if that meant contradicting things they had advocated in the past.

The first step was to critically rethink the nature of the relationship between customer satisfaction and customer spending. Beliefs that could not be supported with hard data were discarded. The team then compared what was proved to be true with how customer satisfaction is currently measured and managed. Surprisingly, despite decades of scientific research into customer satisfaction, two serious disconnects between what researchers and managers know to be true and what is actually done were discovered.

Putting It Together

These two seemingly unrelated issues—(1) satisfaction is relative, and (2) rank matters—are actually integrally related to one another. Regarding issue 1 (satisfaction is relative), the underlying problem is that it is not the absolute “score” that a customer assigns that matters—it is a customer's perception of performance relative to competitive alternatives. Regarding issue 2 (rank and market share are linked), rank is a means of gauging relative position vis-à-vis competitive alternatives.

To see how these issues relate to one another, imagine that your brand has two customers: Janet and John (see Figure 2.1). Both Janet and John rate their satisfaction with your brand a 9 on a 10-point scale (where 10 is the highest level of satisfaction). Nearly all managers would consider this a good score. Using the NPS classification system, both Janet and John would be considered promoters.

Janet and John, however, also use two other brands: Brand A and Brand B. On the one hand, Janet rates her level of satisfaction with Brand A a 9 and her satisfaction with Brand B a 10. On the other hand, John rates his level of satisfaction with Brand A a 7 and with Brand B an 8.

Despite the fact that Janet and John both rate your brand a 9, with John your brand is his clear first choice. With Janet, your brand is tied for last. The result of this difference in rank is that John allocates a substantially higher share of his category spending with your brand than Janet does.

This problem happens all the time. For most industries, customers divide their spending in the category among multiple competing brands. But not all brands are equal in satisfying customers. We would naturally expect preferred brands to get a greater share of customers' spending in the category. Relative “ranked” satisfaction levels easily capture these preferences. As a result, they are more strongly related to share of wallet than the absolute score given to each firm. In fact, our research found that simply transforming absolute satisfaction levels to relative ranks tended to explain more than 20 percent of the variation in customers' share of category spending.⁸ This is a remarkable improvement given that absolute satisfaction levels tend to explain only 1 percent of the variation in share of wallet.

Determining Your Rank

Given that rank is far superior to absolute satisfaction levels in linking to share of wallet, the next logical question is, “What's the best way to determine rank?” The seemingly obvious answer—simply to ask customers to rank the firms they use—turns out not to be the best solution. If you simply ask people to rank things from best to worst, they will forcibly assign a distinct rank (first, second, third, etc.) for each item.

The problem is that these distinct ranks are not how customers typically view the brands they use. Although a customer may have a clear favorite brand he uses, chances are he also uses several brands that he views as being the same. Therefore, having customers force brands into distinct ranks does not reflect their actual perception of these brands.

So it is important to make it easy for there to be ties. Researchers have found that a good way to do this is to rescale customers' ratings of the brands they use into ranks. Most important, this process makes it possible to link the data collected in customer surveys with customers' actual purchase behaviors.

Given that ties are important, the issue then becomes how to account for ties. In most of the rankings we see—sporting events, school rankings, and so on—a tie means that everyone who achieved the same level gets to share the highest rank. In other words, it's possible for more than one team to legitimately say, “We're number one!”

When it comes to customers' share of wallet allocations, though, there are no multiple gold medal winners. If two brands are tied for first place with a customer, neither brand gets the rewards that would accrue if one brand were the sole winner. After all, the size of the customer's wallet doesn't change because several of the brands he uses are tied.

Instead, rewards are divided evenly among tied competitors for the places they would have occupied had they not been tied. So if two brands are tied for first place, they would have occupied spots 1 and 2. As a result, their rank would correspond to 1.5. The next potential rank, assuming no other ties, would be 3 (i.e., third place).

The Wallet Allocation Rule and Share: The Evidence

We embarked on an extensive investigation to see if our belief about the relationship between rank and share of wallet was correct. We surveyed more than 5,500 customers across a dozen industries, collecting customers' satisfaction and loyalty ratings across a wide array of questions. In addition, we obtained a purchase history for each person surveyed. We then examined the data to see if any relationship actually existed between the customer metrics we collected and share of wallet.

What we discovered was quite unexpected. Not only was there a relationship between rank and share of wallet, but it followed a clear pattern that can be predicted by two things: a brand's rank among competitors used and the number of brands used (see Figure 2.2). Most important, this prediction could be calculated using a very simple mathematical formula—the Wallet Allocation Rule.

The correlations between share of wallet and the Wallet Allocation Rule across the industries investigated were remarkable (see Figure 2.3). More important, both the customer-level and the firm-level correlations were strong. These findings indicated that it was possible to predict share of wallet with a high degree of accuracy by simply knowing the rank and the number of brands used by the customer.

Rather than celebrate this discovery, however, we needed to be sure we were right. After all, many seemingly great initial discoveries have failed under the real-world demands of business competition.

When we began this investigation, we expected that finding a strong relationship would require a complex mathematical formula filled with Greek symbols. The Wallet Allocation Rule, however, is so simple that it was hard for us to accept that we were the first to discover it (particularly given the thousands of researchers who examine satisfaction data all the time).

Given our skepticism, we insisted upon rigorous testing of our findings. First, we needed to be confident that the Wallet Allocation Rule would work across cultures. We therefore surveyed more than 7,000 customers in eight non–North American countries (covering four continents) about their usage of credit cards. We selected this industry to minimize the likelihood that industry structure and the uniqueness of competitors in the various countries would significantly influence our results. Our investigation found strong correlations between the Wallet Allocation Rule and share of wallet for all countries examined.

Although these results were impressive, we needed to be certain that the Wallet Allocation Rule would reveal consistent results over time and prove to be a useful key performance indicator for managers to track. Specifically, we needed to be certain that changes in Wallet Allocation Rule scores corresponded to changes in share of wallet over time.

That need, however, presented us with a challenge. It was unreasonable to expect large shifts in customer metrics and share of wallet levels just a few months after completing our initial wave of surveys.

Instead, we needed to examine markets in which customers' share of wallet allocations were changing rapidly. This meant something disruptive had to have happened within a market. The difficulty from a research perspective is that we had to know exactly when this disruption would take place to ensure that we could measure share of wallet before and after the event.

To address this problem, we examined markets in which a new retail store was scheduled to open. Clearly, the opening of a new store dramatically disrupts competitive dynamics in a market area, quickly shifting customers' spending patterns.

We studied two different retail markets covering two distinct product categories, before and after the opening of new stores. The results of this test demonstrated a strong link between the Wallet Allocation Rule and share of wallet regardless of changes in market dynamics and corresponding shifts in customers' share of category spending.

We also went back to five of the eight countries examined regarding credit card usage after approximately six months. The results between the two waves of data were essentially identical, all demonstrating strong correlations.

The most important test had yet to come, though. We needed to know if changes in an individual customer's share of wallet matched changes predicted by the Wallet Allocation Rule. To do this, we had to do something rarely done in customer satisfaction research. Approximately one year after our initial investigation, we went back to the same customers to find out.

The results unambiguously demonstrate that the Wallet Allocation Rule links strongly to individual customer behavior (see Figure 2.4). By comparison, changes in other commonly used metrics show a very weak correlation to changes in share of wallet.

The findings of this research were published in the Harvard Business Review.⁹ One month later, the research received the Next Gen Market Research Award for Disruptive Innovation.¹⁰

Next we sought to replicate these findings through a large-scale study of the U.S. credit union and retail banking market.¹¹ In a survey of 4,712 banking customers across the country, we found that the Wallet Allocation Rule explained 55 percent of the variation in customers' share of deposits. (Share of deposits refers to the percentage of deposits that a customer allocates with a particular financial institution.) By contrast, common metrics such as satisfaction and NPS level explained less than 10 percent. The findings of this study were published in the International Journal of Bank Marketing, a peer-reviewed academic journal in the field of financial services marketing.¹²

Finally, we sought to replicate our findings using a large-scale, multicountry database and a team of leading academic researchers from Northwestern, Vanderbilt, Fordham, and Ghent Universities. We examined 79,543 customers who provided 258,743 satisfaction ratings regarding the brands they use within a particular industry covering more than 650 brands from 22 industries and in 15 countries.

In this investigation, we conducted a comprehensive comparison of the Wallet Allocation Rule and multiple alternative approaches that have either been proposed by other researchers or represent logical choices for comparison based on prior scientific studies. The models were examined using multiple performance criteria. Again, the Wallet Allocation Rule was found to perform as well as other, more complex models in linking to share of wallet. In fact, the absolute correlation between a change in the Wallet Allocation Rule score over time and a change in share of wallet was nominally the largest overall.¹³ The findings of this investigation were published in the Journal of Service Management, a peer-reviewed academic journal in the field of service management.¹⁴

Other researchers have also investigated the Wallet Allocation Rule and found similar results. In one of the most comprehensive investigations, researchers Alice Louw and Jan Hofmeyr compared correlations between the Wallet Allocation Rule and two more complex approaches with customers' actual share of category spending in three industries.¹⁵ Although the survey questions used were not the same across the three approaches investigated, the findings were. The Wallet Allocation Rule worked as well as these more complex approaches.¹⁶

Most new approaches rely on anecdotes to support their claims (e.g., “Firm X adopted this new approach, and it transformed its business”). Although it is always nice to have a story, anecdotes mean something only if they are proved to work across companies and industries.

As these different investigations make clear, the Wallet Allocation Rule has undergone numerous, rigorous scientific tests. More important, it has passed them all!

The “Best” Metric?

The Wallet Allocation Rule makes it possible for managers to easily link their customer metrics to share of wallet. But in a world of many competing metrics (satisfaction, NPS, etc.) what is the best metric to track?

In recent years, researchers have advanced a number of customer metrics purported to be the best at linking customer survey data to business outcomes. But the best metrics have shown only modest correlations to growth. None have shown themselves to be universally effective across all competitive environments. And all of these metrics have proved to be weak predictors of share of wallet. Nevertheless, the weaknesses of existing metrics have not discouraged managers from adopting new ones in hopes of gaining better insight into customer behavior.

In picking a best metric, it is important to recognize that the most commonly used metrics are highly correlated with one another. In fact, statistical analysis of these metrics reveals that they are actually measuring the same construct. So although it is possible that one metric would perform somewhat better than another in some situations, the likelihood that any one metric would be an obvious winner seems highly unlikely.

A key implication of the Wallet Allocation Rule is that the primary problem in linking these measures to share of wallet is in the use of absolute scores, not in the choice of metric. To determine whether this was indeed the case, we examined the linkage of the most commonly used satisfaction/loyalty metrics to share of wallet when using the Wallet Allocation Rule.

Although we expected all metrics to perform reasonably well when using the Wallet Allocation Rule, we did not expect the almost identical results we found (see Figure 2.5). Our findings indicate that efforts to uncover the best metric are misplaced. Certainly, there will be isolated cases in which one metric outperforms another. However, it appears that managers in most industries do not need to switch the customer metrics (satisfaction, NPS, etc.) that they have been tracking over time if they wish to have a strong linkage to share of wallet. Rather, they simply need to apply the Wallet Allocation Rule.

Why Does the Wallet Allocation Rule Work?

It is intuitive that the amount a customer spends on a brand would be a function of how that person ranked that brand vis-à-vis other competitors also used. We would naturally expect the preferred choice to be used more than the next best choice.

But the foundation of the Wallet Allocation Rule goes beyond intuition. The crux of the rule states that share of wallet is a function of a brand's rank. This relationship reflects a scientific empirical law, referred to as Zipf's law.¹⁷

In the 1930s, Harvard linguist George Kingsley Zipf discovered that the frequency of any word in a language is inversely proportional to its rank. Since then, scientists have found similar relationships for a wide variety of situations, including Internet usage, world income distribution, population ranks of cities, the size and frequency of earthquakes, and note usage in musical compositions.

Most important for managers, academic researchers have known for some time that the market shares of companies follow this law. As a result, the most widely proposed alternatives to the Wallet Allocation Rule to link to share of wallet use the Zipf distribution (a mathematical formula based on Zipf's law).

Although Zipf's law is conceptually simple, most managers would find it difficult to calculate the Zipf distribution. Mathematically, the formula for calculating share of wallet using the Zipf distribution is as follows:

where is a constant that depends on the number of brands .

Finding the optimal Zipf distribution requires inputting both rank and share of wallet data into a software application and then computationally back-solving to determine the appropriate values of s in the preceding equation.

This is clearly complicated. Worse still, the model is difficult to convey to senior managers and to the organization at large. The reality is that managers tend to reject overly complicated models because they intuitively don't understand them, can't communicate them, and certainly can't rally the organization around them. As a result, they “revert to models of great simplicity.”¹⁸

Unfortunately, the ultrasimple models (e.g., absolute satisfaction and NPS) do not link to the share of spending that customers give to the brands they use. They may be simple, but no manager will knowingly accept wrong. And if the goal is to drive share of wallet, they are most definitely the wrong metrics to use.

With the Wallet Allocation Rule, managers can easily link their customer metrics to share of wallet without requiring a PhD in statistics. All you need to know is the number of brands that customers use and their rank.

Using the Wallet Allocation Rule

Although the Wallet Allocation Rule isn't as simple as calculating your firm's NPS, it isn't difficult either. To quote the Harvard Business Review, “Don't let the math scare you.”¹⁹ Using the Wallet Allocation Rule is a very simple three-step process.

1. Step 1: Survey customers to determine the brands (or stores or firms) they regularly use in the product category you want to analyze. Let's say that John, Jane, Mary, and Tom all use the same brands of detergent: Brands X, Y, and Z.
2. Step 2: Gauge satisfaction (or another common loyalty metric, such as NPS) for each brand the customer uses; then convert those scores into ranks. The highest-scoring brand for a customer would be ranked first, the second-highest second, and so forth.

Figure 2.6 shows the satisfaction ratings for three brands of detergent for John, Jane, Mary, and Tom.

Figure 2.7 shows the ranks of the three detergents based on the satisfaction scores provided by John, Jane, Mary, and Tom.

In the case of a tie, as was the case for Tom with Brand Y and Brand Z, assign each brand a rank using the average of the two spots they would have occupied had they not been tied.

Brands not used are treated as missing and are not assigned a rank, as was the case for Jane with Brand Y.

1. Step 3: To arrive at a brand's share of wallet for a given customer, plug the brand's rank and the number of brands used by the customer into the Wallet Allocation Rule formula:

For example, John's share for Brand Z is as follows:

Repeat the calculation for each customer and brand. To obtain the average share of wallet your customers give to a particular brand, simply average individual customers' share-of-wallet scores. There are two ways to do this (see Figure 2.8).

• All customers: This is the average share of wallet going to a brand from all of your customers. This provides a quick look at the overall financial opportunity represented by a specific competitor.
• Brand users only: This is the average share of wallet going to a brand from your customers who use this brand (i.e., noncustomers of the brand are excluded). This provides insight into the strength of the relationship that your customers who use a competing brand feel toward that brand.

Wallet Allocation Rule Strategy

Several important strategic issues stem directly from the Wallet Allocation Rule.

• Strategic issue 1: First and foremost, managers cannot evaluate their firms without taking competition into account. Although this sounds obvious, the reality is that managers typically evaluate their performance using customer perceptions of their firm only. As a result, the target objectives used to evaluate and compensate managers are almost never based on changing the perceived rank of the firm vis-à-vis competition. Rather, they are based on achieving a particular score for the firm.
• Strategic issue 2: Rank matters! Every manager knows that it is better to be number one than number two. But the Wallet Allocation Rule makes it very easy for managers to determine the financial implications of that. The difference between first and second is typically quite large (see Figure 2.9). And making that jump can have a tremendous financial impact.
• Strategic issue 3: Parity hurts! The Wallet Allocation Rule also makes clear that it is not enough for your brand to be tied for first place. There must be a reason for customers to prefer your firm. Otherwise, you evenly divide your customers' share of wallet with your closest competitors (see Figure 2.10).
• Strategic issue 4: The more brands a customer uses, the lower the potential for everyone. The number of competitors that your customers use has a significant impact on customers' share of wallet (see Figure 2.11). Ranking first in a field of three is much better than ranking first in a field of six. That's because every brand used by a customer gets some percentage of his or her wallet.

These strategic issues have practical implications for how we identify opportunities for improving share of wallet. The traditional approach to pinpointing opportunities involves answering the question, “What can we do to make you happier?” Whether you're analyzing customers' open-ended survey responses or deriving importance through statistical analysis, the focus is almost always on improving satisfaction with what the firm or brand currently offers.

Performance, however, is relative to competitive alternatives. Obviously, improving satisfaction is important. This is primarily because at some point, increases in satisfaction make a brand more attractive to customers relative to competitors. But simply striving for improved satisfaction is not enough.

Managers also need to understand exactly why customers use each of the brands that they do. Customers have legitimate reasons for using multiple brands in a category. Therefore, efforts designed to improve share of wallet that do not address precisely why your customers also use your competitors are doomed.

Managers can gather this information as part of the Wallet Allocation Rule survey process. And this process doesn't have to be complex. It can be as simple as asking customers something like the following:

1. When choosing between brands, what tends to be the deciding factor in choosing one over the other?
   1. I choose [Brand 1] when…
   2. I choose [Brand 2] when…

Armed with an understanding of why your customers use both your brand and competitive brands, managers can identify what it really takes to be the first choice of their customers. And because the Wallet Allocation Rule is tied to share of wallet, managers can prioritize their efforts by their potential impact on future revenues.

How to Improve Your Rank

One of the key takeaways of the Wallet Allocation Rule is that if you want to improve your share, you need to improve your rank. Improving rank, however, is not the same as improving your overall satisfaction or NPS level. Improving your rank requires minimizing the reasons customers feel the need to use the competition.

Following is an easy-to-follow process you can implement right away to improve your share:

1. Survey a statistically valid sample of your customers to find out how they rank you and the competitors they also purchase from.
2. Apply the Wallet Allocation Rule to establish the share of wallet of each competitor.
3. Determine how many of your customers use each competitor.
4. Calculate the revenue that goes from your customers to each competitor.
5. Identify the primary reasons your customers use your competitors.
6. Prioritize your opportunities to improve your share of wallet: Estimate the cost of addressing each reason your customers choose a competitor and weigh those costs against your potential financial return in each case. Remember to take into account the cumulative impact of addressing issues that apply to multiple competitors.

The Rule in Practice

To see how this would work in the real world, let's look at a case drawn from our research. Management at a grocery retail chain surveys its customers and finds that most customers are happy with their experience; 53 percent of customers would be classified as promoters using the NPS classification system (in the “would recommend” category, they rate the store a 9 or 10 on a scale of 0 to 10).

Despite these good scores, however, only 43 percent of customers would classify the grocer as their first choice. In other words, 57 percent either prefer one or more of the grocer's competitors or consider the grocer to be at parity with them.

To understand the financial implications, the grocer used the Wallet Allocation Rule to calculate how much of its customers' spending is going to its competitors. To do this, the grocer first calculated the average share of wallet it gets from its customers and the portion its competitors get from its customers.

Next, because share of wallet reflects the percentage of spending customers give to a brand, the grocer multiplied share of wallet estimates by its customers' average monthly grocery spend and the number of its customers who also patronize the competing stores. The results indicated that its top three competitors were extracting $425 million from its customers' wallets.

To understand what could be done to get some of this money back, the grocer analyzed its survey data to understand what was driving customers' spending behaviors.

The company conducted an analysis to determine the most important drivers of its NPS. This analysis revealed that the top two reasons customers recommend the grocer are the quality of its produce and the ambience of the store. This was no surprise to management, as the grocer competes by positioning itself on providing superior produce and maintaining a quality store atmosphere.

An analysis of the drivers of rank, however, showed that improving these attributes would not result in customers giving the grocer a greater share of their spending on groceries. The analysis revealed that the grocer's customers are using the competition for completely different reasons.

For one competitor, the primary attraction is everyday low prices. Another competitor also competes on price, but largely through rotating deep discounts. And a third competitor's main appeal is the convenience of its locations—it had large numbers of smaller stores distributed throughout metropolitan areas.

For the grocer to move up to first place in more of its customers' minds, it can't simply do more of what it already does well; providing even better produce or enhancing the aesthetics might further delight customers who already rank it first but would be unlikely to change the minds of the rest, who are mainly interested in low prices and convenience.

The analysis made clear that, at least in the short term, focusing on the convenience-based competitor was not a realistic option. The grocer certainly wasn't going to be able to open lots of new stores to have comparable levels of convenience.

This meant that the grocer was forced to focus on its price-based competitors if it wanted to get a greater share of its customers' wallets. Of course, a high-quality ambience, high-quality produce store is never going to be the price leader—its costs of operations are naturally higher.

Matching price, however, is not the goal. The goal is to reduce customers' need to use the competition.

To do this, the grocer determined what customers were actually buying from the competition. Because the company had good information on the purchases its customers made, it could easily see what its customers were purchasing. More important, it could see what they were not purchasing but would be expected to purchase as part of a total grocery shopping basket. It also confirmed the items that were going to competitors through customer surveys.

These missed purchases tended to be staple goods, such as bread, milk, sugar, and flour. Because staples tend to offer little differentiation, they largely compete on price. So the fact that many of the grocer's customers choose to purchase staples from the cheaper competitors resonated with management.

This also presented management with a problem. There was no way that the grocer could profitably match the lowest price on most staples.

Instead, the grocer reasoned that it should take advantage of the fact that its customers are already attracted to the store for its produce and ambience. Because customers are already visiting the grocer, management's goal should be to make it less economically attractive for customers to shop multiple grocery stores. This, however, doesn't require being the price leader. Instead of matching the cheapest competitor, the grocer lowered the price on its most commonly purchased staples to a relevant level—low enough that more customers perceived it was better to consolidate their grocery purchases into one shopping trip.

Surveys after the price change found that 49 percent of customers considered the grocer their first choice (a gain of 6 percent). Moreover, the average number of stores customers regularly shopped dropped from 2.5 to 2.0. These changes, when plugged into the Wallet Allocation Rule, translate to a 7-point increase in share of wallet—the equivalent of shifting $62 million from competitors' registers into the grocer's own.

Conclusion

Many companies could see this kind of revenue jump if they decided to stop pursuing customer satisfaction for its own sake and instead focused on how satisfaction and other loyalty boosters could help them pull ahead of the competition.

The Wallet Allocation Rule turns traditional satisfaction measurement on its head. The rule shifts the emphasis from internally focused satisfaction measurement to your brand's competitive position in the marketplace.

Brands exist in the market, not in a vacuum, and that's the way to approach performance. Sounds elementary, right? Yet most managers treat satisfaction and loyalty metrics as if just achieving a particular score is sufficient. The reality is that simply boosting your brand's satisfaction ratings rarely increases your share of wallet—but improving your brand's rank does.

The Wallet Allocation Rule allows you to build strategies that directly affect brand performance and then measure their impact on share of wallet.

What do managers normally do to try to improve share of wallet? Typically, they create programs and initiatives designed to improve customer happiness and then measure success based on satisfaction. Increased customer happiness is important, but it rarely has the desired effect on the bottom line.

Managers should instead focus their energy on understanding why their consumers use the brands they do. If your brand is not their top choice, find out why they prefer your competitor's brand over yours.

By applying the Wallet Allocation Rule, managers get real insight into the money they currently get from their customers, the money available to be earned from them, and what it takes to get it.

If growing your share is what you're after, you won't learn much from watching changes in your satisfaction scores. Focus instead on how to pull ahead of your competition. That's what makes a champion.