Long customer relationships can be low-profit

A long customer relationship is a duration record. It is not, by itself, a profitability record.

The short answer is that longevity should be reviewed alongside revenue, service cost, price, margin, measurement quality, and the investment decision. A relationship can last while its economics change. A customer can purchase repeatedly while requiring more service, receiving lower prices, or generating less margin than the retention slogan suggests.

Reinartz and Kumar test whether long customer relationships in a noncontractual setting are associated with higher lifetime profitability, increasing profits over time, lower service costs, and higher prices. Their evidence shows that noncontractual customer longevity does not automatically imply greater profitability because customer revenue, costs, and pricing can evolve differently over time. Mulhern provides a conceptual and methodological foundation for measuring customer profitability, extending customer-lifetime-value approaches to broader target-marketing applications. He emphasizes explicit inputs and the concentration of profits across customers. Together, the sources support an economic diagnostic, not a universal retention rule.

Why must customer profitability analysis isolate the economic transaction object?

Imagine a customer relationship whose duration is clearly visible. The account has purchased for several periods. What is not yet visible is whether the relationship is becoming more or less valuable. Revenue may be stable while service cost rises. Price may be discounted. A large account may contribute a large share of revenue but not a large share of profit.

Before calling a relationship healthy, name these fields:

Field	Question	Why it matters
Relationship clock	How is duration measured, and from which event?	Longevity is a time variable whose start and end rules affect comparison.
Revenue path	What revenue is observed over which periods?	Repeated revenue does not reveal cost or margin by itself.
Service cost	What support, delivery, acquisition, and account costs are included?	Cost-to-serve can move differently from revenue.
Price path	How do discounts, terms, and realized prices change?	Price can explain why a long relationship is or is not profitable.
Margin boundary	Which contribution or profit measure is being used?	The conclusion depends on the economic definition.
Evidence quality	Which inputs are measured, allocated, or estimated?	Unspecified inputs make profitability comparisons fragile.
Investment decision	What action would change if the economics changed?	A diagnostic needs a decision, not only a segment label.

These fields keep retention, longevity, and profitability separate. Retention asks whether the relationship continues. Longevity records how long it has continued. Profitability asks what remains after the relevant costs and prices are included. The three can correlate in one setting and diverge in another.

Review field	Required input	Permitted interpretation	Stop signal
Relationship clock	Duration rule and starting event	“Longevity is measured on this clock.”	A longer relationship is assumed to be better.
Revenue trajectory	Period revenue and purchase pattern	“Revenue follows this path.”	Revenue is used as a profit proxy without qualification.
Cost and price	Service cost, discounts, terms, and realized price	“The cost and price boundary is explicit.”	Cost-to-serve and price changes are omitted.
Margin boundary	Contribution or profit definition	“Profitability means this measure.”	A lifetime-value label hides the calculation.
Evidence quality	Measured, allocated, or estimated inputs	“The result carries this measurement uncertainty.”	Unspecified allocations are treated as facts.
Investment decision	Retain, serve, develop, or review	“This action follows this economic question.”	The segment is ranked without a decision rule.

Why does extended account tenure fail to produce superior profit margins?

Reinartz and Kumar test the marketing assumption that long-life customers should always receive disproportionate attention. Their setting is noncontractual, which matters because the relationship can end without a formal cancellation event. The study examines whether long customer relationships are associated with higher lifetime profitability, increasing profits over time, lower service costs, and higher p

rices.

The reported conclusion is a useful correction to the retention slogan: noncontractual customer longevity does not automatically imply greater profitability. Customer revenue, costs, and pricing can evolve differently over time. A relationship can be long and still have an unattractive margin if service demands increase, prices fall, or future purchases are overestimated.

This does not imply that long relationships are bad. It means that duration is not a sufficient decision criterion. The source evaluates an assumption against evidence in a noncontractual setting. It does not say that every long-life customer is unprofitable, and it does not establish the same result for every contractual model, category, or service design.

The permitted statement is therefore bounded: in the reported noncontractual setting, longevity did not automatically imply greater profitability because revenue, cost, and pricing followed different paths. The stronger statement, “long relationships are low-profit,” is as unsupported as “long relationships are always high-profit.”

How do account servicing inputs determine true net operating margin?

Mulhern provides a conceptual and methodological foundation for measuring customer profitability by extending customer-lifetime-value approaches to broader target-marketing applications. The contribution is a reminder that a profitability number is made from inputs. If the inputs are not specified, the number cannot carry a clear economic interpretation.

Customer-profitability analysis depends on explicitly specifying inputs and measuring how concentrated profits are across customers. That concentration question changes the managerial conversation. A portfolio may have many customers with positive revenue and a smaller group contributing most of the profit. A long relationship can sit in either group. The account’s duration does not identify its contribution without the relevant revenue, cost, price, and allocation rules.

Mulhern also combines a business-to-business customer-profitability analysis with propositions about determinants of profitability and a future research agenda. Those propositions are not a ready-made score for a private account. They are a reason to make the economic object explicit before a portfolio is ranked.

The practical rule is simple: write the calculation boundary before debating the retention action. Specify which service, acquisition, delivery, discount, and account-management inputs are included or excluded. If an input is allocated rather than directly measured, label it as an allocation. A clean-looking number with an unclear boundary is not stronger evidence.

How do revenue, servicing costs, realized price, and capital diverge over time?

A relationship-profitability review should track several paths at once:

1. Revenue: what was bought, in which periods, and at what realized value?
2. Service cost: how much support, delivery, customization, and account attention did the relationship require?
3. Price: which terms, discounts, or concessions changed the realized price?
4. Margin: which contribution measure remains after the chosen cost boundary?
5. Investment: what additional resources are being considered, and what return would justify them?

The list is author synthesis, but it follows directly from the sources’ separation of longevity from economic inputs and the requirement to specify customer-profitability measures. It also prevents a common substitution: using retention as a proxy for future profit and then using that proxy to justify more investment.

The same logic applies in the other direction. A short relationship is not automatically unattractive. It may have high margin, low service cost, or strong future potential under the chosen boundary. The evidence does not support a universal preference for either long or short relationships. It supports more precise measurement.

What does a synthetic account model reveal about high-volume margin erosion?

Take a synthetic account with a long relationship clock and stable annual revenue. The account receives a discount, requires frequent support, and has a customized delivery process. A second account has a shorter clock, less revenue, and a standard service path. Without a cost and price boundary, the first looks better. After the boundary is written, the ranking may change. That is not a surprising data trick. It is the result of asking a different economic question.

The example still leaves evidence gaps. Are support hours measured or allocated? Is the discount recorded at the transaction level? Are future purchases observed or projected? Is the margin definition consistent across accounts? Does the investment decision concern retention, expansion, or service redesign? The relationship card keeps those questions visible without inventing a portfolio conclusion.

The result is a more honest decision statement: “Under this cost, price, and margin boundary, the available inputs support this investment review.” It is not “long customers are worth more.”

What disciplined governance sequence should guide relationship contract renegotiation?

Use this sequence before a claim about customer longevity or profitability enters a portfolio decision:

1. Define the relationship clock and whether the setting is contractual or noncontractual.
2. Name the revenue, service-cost, price, and margin inputs.
3. Identify which inputs are measured, allocated, estimated, or missing.
4. Measure or describe how profits are concentrated across customers.
5. Separate retention, longevity, current profitability, and expected future value.
6. State the investment action and the threshold that would change it.
7. Test whether the conclusion changes under plausible cost, price, or allocation assumptions.
8. Keep the source setting visible before generalizing to another business model.

This sequence does not produce a universal customer score. It produces a decision record that can be reviewed when the inputs or the relationship change.

Where are the empirical limits of customer relationship profitability models?

The two sources do not analyze private accounts or recommend an individual retention action. Reinartz and Kumar’s result is bounded to a noncontractual setting and does not establish a universal rule for contractual relationships. Mulhern provides a measurement foundation and propositions, not a ready-made profitability ranking for a private portfolio.

The stopping rule is concrete. Do not release “longer relationships are more profitable” or “longer relationships are low-profit” until the relationship clock, revenue, service cost, price, margin boundary, measurement quality, and investment rule are named. If one is missing, longevity may still be interesting. It is not yet a defensible profitability claim.

The relationship-profit boundary connects to customer lifetime value as a forecast and retention ranked by profit.

References

1. Mulhern, F. J. (1999). Customer Profitability Analysis: Measurement, Concentration, and Research Directions. Journal of Interactive Marketing, 13(1), 25-40. DOI
2. Reinartz, W. J., and V. Kumar. (2000). On the Profitability of Long-Life Customers in a Noncontractual Setting: An Empirical Investigation and Implications for Marketing. Journal of Marketing, 64(4), 17-35. DOI