Most businesses believe they know their best customers. A restaurant owner can name the regulars. A golf course knows its long-standing members. A venue knows which organizations book large events. Ask who matters most to the business and there are usually a few customers—or at least a type of customer—that immediately comes to mind.
Turning that intuition into analysis is surprisingly difficult.
Consider a golf course trying to identify its best customers. Its membership and reservation data show that golfers between 45 and 64 make up the largest portion of its customer base. Higher-income households are well represented, and many customers work in professional or management occupations.
That is a useful description of the audience. It does not identify the course’s best customers.
The customers who play most frequently could belong to a different group. The highest annual spending might come from members who play less often but purchase lessons, food and beverage, merchandise and other services. Corporate outings might produce substantial revenue from people who otherwise appear only once or twice a year. Some members may generate unremarkable revenue in any individual year but remain with the course for a decade.
All of these customers can be valuable. They are valuable in different ways.
That is the problem hidden inside the phrase “best customer.” Before customer data can identify one, the business has to decide what outcome it is trying to evaluate.
Customer value can take several forms.
Economic value may appear through revenue, contribution margin or lifetime economic value.
Relationship value may appear through retention, tenure or sustained recurring engagement.
Utilization value may come from frequent use or from customers who make productive use of otherwise available capacity.
Strategic value can reflect a customer group’s importance to a market, service or direction the business is intentionally trying to develop.
These forms of value can overlap, but they should not automatically be treated as substitutes for one another. A long-tenured customer may also be highly profitable. A frequent customer may also generate substantial revenue. When the evidence establishes both, the business can say so. When it establishes only one, the distinction matters.
For some businesses, annual revenue is the obvious place to start. For others, frequency, retention, lifetime value or profitability may matter more. A business with excess capacity may place unusual value on customers who use otherwise quiet periods. Another business operating near capacity may care much more about the revenue or margin generated from each scarce unit of capacity.
Even two golf courses serving similar markets can therefore have very different definitions of a desirable customer. A course trying to fill weekday tee times may place particular strategic value on golfers who can play regularly during those periods. A course already operating near capacity may be more interested in customers who participate in outings, instruction, dining or other services in addition to golf.
The customer didn’t change. The business objective did.
“Best customer” is not a demographic category. It is a business outcome.
That distinction matters because demographic data is often easier to obtain than economic data. It is relatively straightforward to discover that a large portion of an audience falls within a particular age or income range. It can be considerably harder to determine which customers produce the strongest margins, remain longest, use constrained capacity most efficiently or generate the greatest lifetime value.
The information that is easiest to measure can therefore begin standing in for the information the business actually wants.
A company may know that its most frequent customers are homeowners with relatively high household incomes and gradually begin referring to that profile as its “best customer.” But unless frequency is itself the outcome the company values—or has been shown to correspond with the outcome it values—the analysis has skipped a step.
Frequency measures frequency. Revenue measures revenue. Retention measures retention. Profitability measures profitability.
They may be related, sometimes strongly. They are not interchangeable.
Revenue can be easier to measure than value
Revenue illustrates the problem particularly well.
Suppose two customer groups each generate $500,000 annually. One requires substantial discounts, service time or other variable costs while the other does not. Their revenue contribution is identical; their economic contribution may not be.
The ideal analysis would account for contribution margin, customer acquisition costs, servicing costs, retention and other factors appropriate to the business. Many organizations—particularly smaller ones—do not have customer-level information at that level of detail.
That doesn’t make customer analysis impossible. It does make precision about terminology important.
If the data identifies high-revenue customers, they are high-revenue customers. If it identifies frequent customers, they are frequent customers. If it identifies customers who have remained for many years, they are long-tenured customers.
Calling any of them “most profitable” requires evidence about profitability. Calling them “highest value” requires either a defined measure of value or an explicit business decision about which outcomes matter.
The same applies to engagement. A customer who visits every week clearly has a different relationship with the business from one who visits once a year. That recurring relationship may be extremely important. But a frequent customer can also be highly price-sensitive, purchase little per visit or consume capacity that could have been sold differently.
Meanwhile, a customer appearing only a few times each year could generate substantial revenue every time.
The useful question isn’t whether frequency or tenure is “better” than spending. It is whether any of those measures represent, or are demonstrably associated with, the outcomes the business actually cares about.
Once that outcome is defined, the Audience Profile developed earlier becomes much more interesting.
Imagine that our golf course identifies a group of customers who have remained members for at least five years, play regularly and generate above-average annual revenue. Instead of declaring that group valuable because of its demographics, the course can now work in the opposite direction: first identify customers associated with desirable business outcomes, then examine what distinguishes them.
Perhaps they are disproportionately concentrated in several communities. Certain professional fields may appear more frequently. They may have joined through particular programs or events. They may use other parts of the facility at higher rates.
Or there may be very little demographic difference between them and everybody else.
That would be useful to know too.
Businesses often assume that their strongest customers must correspond to a recognizable demographic profile. Sometimes the more important differences are behavioral or relational: how customers discovered the business, how they use it, how frequently they return, which services they combine and how the relationship changes over time.
A Best Customer Profile built from those relationships can look quite different from a traditional marketing persona.
It may also reveal more than one kind of valuable customer.
A golf course could have long-tenured members who provide stable recurring revenue, corporate customers who produce high-value events, destination golfers who purchase premium packages and local frequent players who fill otherwise underused tee times. Trying to collapse those relationships into one fictional “ideal golfer” would remove much of what makes the analysis useful.
The better objective is to understand which customer relationships create value, what kind of value each creates and what evidence supports that conclusion.
A pattern is not an explanation
Once a business begins comparing high-value customer groups with its broader audience, interesting patterns will inevitably appear. This is also where analysis can begin moving too quickly.
Suppose the course finds that many of its long-tenured, high-spending members live in three nearby communities. Those communities may deserve attention, but the concentration doesn’t explain itself.
Perhaps the drive is particularly convenient. Perhaps members have recruited friends and neighbors over many years. A competing course may have closed. Historical marketing may have been concentrated there. The communities may share demographic characteristics associated with the course’s customer base.
Several explanations could be operating at once.
The same caution applies when demographic characteristics appear disproportionately among valuable customers. If homeowners are heavily represented, homeownership didn’t necessarily cause those customers to become valuable. It may be associated with age, income, geography, household stage or another factor entirely.
The objective is not to create a causal story around every pattern. It is to identify characteristics that appear unusually often among customers associated with important outcomes and decide which patterns deserve further investigation.
This is an important difference between a useful Best Customer Profile and the kind of customer persona that can become detached from evidence. The profile should describe what distinguishes customers associated with value. It shouldn’t invent motivations to make those distinctions more interesting.
There is also a temporal problem. The customers who have been most valuable historically aren’t necessarily the customers a business should spend the next decade trying to reproduce.
A company’s strongest historical customer group may be aging or shrinking. A new service may be intended for customers who currently account for very little revenue precisely because the business has never offered them much. A company may intentionally be moving toward a different market, price point or business model.
Historical customer data tells a business who created value under the conditions that existed when the data was generated.
Strategy asks where the business wants value to come from next.
The two deserve to be compared, not confused. A strategically important customer group is a target for future development; it should not be presented as a historically demonstrated Best Customer group unless the outcome evidence supports that conclusion.
This is why the most useful Best Customer analysis may not produce a single answer. It may identify a stable core of historically valuable customers, several different forms of demonstrated customer value, and one or more strategically important groups the business wants to develop.
That is less tidy than naming one ideal customer.
It is also much closer to how businesses actually work.
From understanding value to finding opportunity
The real advantage of defining best customers carefully appears when the business starts thinking about growth.
If the course knows only that its typical customer is a 45-to-64-year-old homeowner with relatively high household income, it can search for communities containing lots of people with those characteristics. That may produce a long list of places that look demographically attractive while having little connection to the reasons customers actually choose the course.
If instead the business has identified customer groups associated with meaningful outcomes, it can ask a more useful set of questions.
Where do those customers come from? Are they unusually concentrated in particular communities? What characteristics distinguish those places? Are there similar markets where the course attracts surprisingly few customers? Are there customer groups with strong economics but relatively low penetration?
Those questions still don’t produce automatic growth recommendations. A community can resemble an existing strong market and fail for reasons the data hasn’t captured. Nor does a Best Customer Profile prove that people sharing the same characteristics elsewhere will produce the same outcomes.
But the search is no longer based on an imaginary average customer.
It begins with demonstrated business relationships and uses the characteristics associated with those relationships as evidence for where further opportunity may be worth investigating.
That is ultimately what makes a Best Customer Profile useful. It isn’t a description of the person a business would most like to see walk through the door. It is an evidence-based attempt to understand which existing customer relationships are associated with the outcomes the business values—and why those relationships deserve closer attention.
Only after that work does the familiar growth question become meaningful:
Where can we find more customers like our best ones?