Sales performance shows how much a customer has purchased from a company, but it does not indicate what share of the total category budget the company has captured. Share of wallet research helps fill this gap by combining customer declarations, transaction data, and information on purchasing behavior within a given category.
How to measure share of wallet and what is a customer’s share of wallet?
Share of wallet refers to the portion of a customer’s spending in a specific category that goes to the brand, supplier, or company being studied. In its simplest form, the metric is calculated by dividing the value of purchases from a given company by the customer’s total purchase value in the same category and during the same period.
Three elements of this definition are key: a properly defined category, a clearly specified period, and a consistent unit of measurement. If a customer purchases products across several similar segments, it is necessary to determine before the study begins whether the denominator should include spending on the entire category, a selected product type, or only purchases related to a specific use case. The category will be defined differently for a company supplying production materials, differently for a telecommunications service provider, and differently again for a fast-moving consumer goods brand.
The metric can be measured by value, volume, or share of orders. Value-based measurement is useful when purchases differ in price, contract size, or scope of service. Volume-based measurement works well when units are comparable, for example, packages, individual items, or licenses. Share of orders, in turn, shows how often a supplier is selected, but may not reflect the actual value of purchases.
The basic formula follows this logic: a customer’s spending with the company being studied is divided by that customer’s total spending in the category being analyzed. If a customer spent a certain amount with the company while also purchasing similar products or services from other suppliers, only knowing both values makes it possible to calculate the customer’s share of wallet. CRM data alone usually shows the numerator of this calculation, but not the full denominator.
For this reason, share of wallet research is particularly useful when there is a need to understand the purchasing relationship beyond an organization’s own transaction database. Internal data makes it possible to observe order history, average basket value, or contract activity. However, it does not reveal whether a decline in purchases means reduced spending across the entire category, a change in purchasing behavior, or a shift of part of the budget to other companies.
In B2C research, the respondent is typically a person who makes or co-makes purchasing decisions. In B2B customer research, the situation is more complex because purchasing may involve the procurement department, the product user, the person responsible for the budget, and the decision-maker approving the contract. Before measurement begins, it is therefore necessary to determine who knows the actual category spend and who can attribute it to suppliers.
Share of wallet is not the same as declared brand affinity. A customer may evaluate a company positively, use it regularly, and at the same time allocate only a small share of their category purchases to it. For this reason, the metric should be interpreted alongside information on purchase frequency, the types of products purchased, relationship length, and purchasing decision-making processes.
How to design share of wallet research and combine it with sales data?
Reliable share of wallet research begins with defining the population, reference period, and spending category. The period should reflect the natural purchasing cycle. For frequently purchased products, respondents may be able to recall spending over a shorter period. For contract-based, seasonal, or project-based purchases, a longer period may be needed to cover the full decision-making and delivery cycle.
The next step is to select data sources. Most often, a company’s own data, including sales data, is combined with a survey or in-depth interviews. Sales data provides information on the actual value of purchases made from the company being studied. A survey makes it possible to estimate the total category budget, other suppliers, purchasing structure, and circumstances that affect changes in spending.
Before developing the question list, it is worth determining what information needs to be obtained from the customer in order to calculate the metric and assess the quality of declarations. The questionnaire may include questions about:
- the total value of spending in the category during the specified period,
- the value or share of purchases allocated to the company being studied,
- other suppliers or brands used,
- order frequency and typical purchase volumes,
- the range of products or services purchased from each supplier,
- the respondent’s role in the purchasing process and their level of confidence in the answers,
- changes in the category budget that may have affected comparability between periods.
Spending can be collected directly by asking for an amount, or indirectly by asking about ranges, volume, and average price. A direct question is straightforward to analyze, but may be difficult for respondents who do not know the full budget. Indirect questions increase accuracy in some categories, but require a more extensive questionnaire and subsequent calculation of the responses.
In B2B customer research, it is good practice to separate questions about routine purchases from project-based purchases. A one-time large order may significantly change a customer’s share of wallet even though it does not reflect the ongoing model of cooperation. The survey can therefore ask separately about current purchases, purchases made as part of projects, and spending resulting from contract renewals.
In share of wallet analysis, a loyal customer with a low share of wallet may represent a greater growth opportunity than a new customer, but this requires confirmation in data on the entire category rather than in CRM data alone. CRM records activity with one company, but does not show the scale of purchases made elsewhere. Such a research conclusion should not be based solely on contact frequency, satisfaction scores, or relationship length.
After the surveys have been collected, the data should be prepared for integration with the sales database. Data can be matched at the level of the customer identifier, company, branch, customer segment, or buying group. In B2B, particular attention should be paid to whether the respondent describes spending for the entire organization, one facility, a specific department, or their own area of responsibility. A mismatch in data levels is one of the more common causes of incorrect interpretation.
Combining sources makes it possible to validate responses. If a respondent declares the value of purchases from the company being studied, it can be compared with the value recorded in sales data. A discrepancy does not always indicate a respondent error. It may result from purchases through an intermediary, a different billing period, purchases made by an affiliated entity, or lack of awareness of the full scope of the business relationship. Any major difference should be flagged and included in the data cleaning procedure.
In addition to the metric itself, customer potential can be analyzed, meaning the difference between estimated total category spending and spending recorded with the company being studied. This is an analytical measure that describes the untapped share of the category budget in relation to a given customer. It does not replace customer value, but complements it with the perspective of spending outside the organization’s own database.
What can distort share of wallet results?
The main limitation of share of wallet research is the quality of data on total customer spending. Respondents do not always remember amounts, do not always have access to financial data, and may understand category boundaries differently. Measurement therefore requires a precise description of what should be included and what should not be included in the declared budget.
In B2C research, purchase recall can be a challenge. A customer may remember purchases from major brands while overlooking small purchases, online orders, or products bought occasionally. Questions that support recall of behavior can help, for example questions about the purchase channel, frequency, most recent purchase, product type, or usage situation. However, questions should not be multiplied unnecessarily, as an excessively long questionnaire increases the risk of mechanical responses.
In B2B customer research, a common mistake is asking one person about purchases for which they are not responsible. An employee may know their own orders but not know what other departments or organizational locations purchase. In such situations, the scope of the question should be limited to the respondent’s area of knowledge, or multi-source sampling should be used, in which information is collected from people performing different roles.
Non-comparable periods are another pitfall. If sales data covers a calendar year while the respondent answers a question about recent months, the calculated share of wallet will be based on different time denominators. A similar issue arises when internal data is recorded by invoice date while the customer reports spending by delivery date or order date. The study design should specify from the outset how the measurement period will be standardized.
It is also important to distinguish share of wallet from customer loyalty. Brand loyalty can be described in terms of attitudes, willingness to repurchase, recommendation, or limiting supplier changes. Share of wallet, by contrast, measures purchasing behavior in relation to the entire category. These measures may be related, but they are not interchangeable.
Small sample sizes also require caution in interpretation. If the analysis includes a small number of customers with highly varied purchase values, the average share may be dominated by individual large entities. In such cases, the distribution of results, median, share ranges, and results for clearly defined groups should be presented instead of relying solely on one average.
The absence of a named competitor should also not automatically be treated as an absence of purchases from other companies. A respondent may not remember the supplier’s name, may purchase through an intermediary, or may not know about decisions made by other people. The questionnaire should allow responses such as “other supplier,” “purchase through a distributor,” and “don’t know,” and these should then be considered in the data quality analysis.
How should share of wallet be interpreted alongside customer value and CRM data?
Share of wallet is most useful as part of a set of metrics rather than as a standalone measure. CRM data shows the history of the relationship with an organization’s own company, while share of wallet describes that relationship against total category spending. Combining both perspectives makes it possible to distinguish a customer with high purchases from the company from a customer whose purchases are also high across the category as a whole.
Customer value can be calculated based on revenue, margin, purchase frequency, and relationship length. Share of wallet answers a different question: what portion of the category consists of purchases from the company being studied? A customer with high value in internal data may have either a high or low share of wallet, depending on the size of the total category budget.
When interpreting results, it is worth creating analytical segments that do not combine different phenomena. Share of wallet can be compared with purchase frequency, declared customer loyalty, the range of products purchased, length of cooperation, and the company’s role in the purchasing process. The purpose of such comparisons is to describe patterns in the data, not to automatically attribute causes to changes.
Before presenting the results, it should be clearly stated whether the metric applies to active customers, the entire database, category buyers, or only people with knowledge of spending. Each of these populations provides a different picture. It should also be indicated whether the result is based on declarations, transaction data, or combined sources, and how missing responses were handled.
A well-prepared share of wallet research report should include the category definition, measurement period, calculation method, sample structure, data matching rules, and limitations of interpretation. This enables the audience to understand what the measured share represents, where the data comes from, and what the limits are when comparing individual customer groups.
Frequently asked questions
What is share of wallet?
Share of wallet is the share of a customer’s spending in a given category that goes to a specific brand or supplier. The metric requires knowledge of both purchases from the company being studied and the customer’s total spending in that category. It is not the same as satisfaction or declared loyalty.
How do you calculate a customer’s share of wallet?
To calculate a customer’s share of wallet, divide the value of the customer’s purchases from the company being studied by their total spending in the same category and during the same period. Data on purchases from the company can be obtained from sales systems, while data on the entire category is usually collected through a survey. It is crucial to use the same category definition and period in both sources.
How does share of wallet differ from market share?
Market share describes the portion of total category sales achieved by a company in the market being studied. Share of wallet relates to an individual customer or a group of customers and shows what share of their spending goes to a given company. Market share requires data on the scale of the entire market, whereas share of wallet is based on data about individual purchases.
Ask about a share of wallet study for your customers
Hume’s Institute can prepare a share of wallet measurement based on sales data, a customer survey, and a clearly defined method for combining sources.