A loyalty program can attract thousands of members, distribute millions of points, and report strong customer engagement while doing very little to increase profitable repeat purchases.
The question every marketing leader should ask before approving the budget is simple:
What will this program encourage customers to do that they would not have done anyway?
That distinction matters because customer loyalty and loyalty program performance are not the same thing. A business can reward its most loyal customers generously without changing their purchasing decisions.
The opportunity is to understand which customers need an incentive, which need a better experience, and which would return regardless.
Loyalty Programs Can Look Successful Without Creating Loyalty
Picture the quarterly marketing review.
Your loyalty program has attracted 25,000 members. Enrollment is growing. Customers are accumulating points. Members spend more than nonmembers.
The dashboard looks encouraging.
But what does it prove?
If your most frequent customers were already more likely to join, their higher spending may have little to do with the program.
Consider a simple example. Before joining, your eventual loyalty members were spending $600 annually, while other customers were spending $350. Six months after launch, members still spend more.
You cannot attribute that difference to rewards without accounting for their purchasing behaviour before enrollment.
This is selection bias. The people most inclined to stay may also be the people most inclined to join.
A 2007 study of Dutch grocery shoppers, covering seven loyalty programs and 20 supermarket chains, found that membership had a small positive effect on share of wallet. However, the estimated effect was seven times smaller than a simpler analysis that failed to account for customers selecting themselves into programs. research.rug.nl
That does not mean loyalty programs are ineffective.
A 2022 meta-analysis covering 429 effect sizes across multiple industries found evidence that loyalty programs improve customer loyalty, particularly purchasing behaviour. It also found that results differ according to program design and industry. IDEAS/RePEc
Both findings matter.
Loyalty programs can work. But participation, spending, and points accumulation are insufficient evidence that yours is working.
You need to measure changed behaviour.
The Second Purchase Problem
In a November 2026 MarTech article, Tim Boylan described a client whose loyalty program had a serious mismatch between its reward structure and its customers’ purchasing habits.
Approximately 90% of customers had purchased only once. About 70% of the points issued were too small to redeem. Customers could access rewards only after their fourth purchase.
Most customers never reached that point.
Boylan examined what would happen if the existing loyalty budget focused instead on encouraging second purchases. His analysis projected approximately seven times the earnings of the existing program.
That was a projection from one client case, not a proven outcome or an industry benchmark. But the underlying question deserves attention. MarTech
Why invest in rewarding a fourth transaction when most customers never make a second?
For a business with a large one-time customer population, the first repeat purchase may be the more important problem to investigate.
Perhaps customers need a timely offer after their first purchase. Perhaps they need help using the product, a reminder when replenishment becomes relevant, or an easier reorder process.
A reward that arrives after four transactions has limited influence on a customer who cannot see a reason to make transaction two.
Research supports paying attention to when and how incentives influence behaviour.
A 2025 Keio University working paper examined a randomized coupon experiment involving nearly 130,000 supermarket customers in Japan. The researchers estimated that coupon redemption reduced churn probability by approximately 12 percentage points and increased customer lifetime value for the promoted product by more than 30%. Importantly, benefits at the manufacturer level did not necessarily translate into equivalent benefits for the retailer. Institute for Economic Studies, Keio University
The lesson is not that every business should offer a second-purchase coupon.
It is that the timing, target customer, product economics, and beneficiary of an incentive all matter.
A company selling weekly consumables has a very different repeat-purchase opportunity from a business selling equipment that customers replace every five years.
Before designing rewards, understand the natural purchasing cycle. Then determine whether a program can influence it profitably.
Are You Solving the Wrong Customer Problem?
A customer makes a first purchase and never returns.
Marketing assumes an incentive might help.
But what if that customer had trouble reaching support? What if the product arrived late, onboarding was confusing, or reordering required another lengthy conversation?
A discount does not solve those problems.
This is where customer retention becomes an operational issue.
Gartner’s 2024 research on customer service effort reported that only 14% of customers experiencing a high-effort service interaction said they were likely to continue doing business with the company. This measures stated intentions, not observed repeat purchases, but it shows why difficult service experiences deserve attention. Gartner
A 2024 academic paper in Psychology & Marketing adds an important qualification. Not every form of customer effort is harmful. Some effort contributes to learning, customization, or customer value. Businesses should focus on eliminating frustrating, unnecessary friction rather than removing every interaction. Wiley Online Library
For a marketing leader, the practical distinction is straightforward.
If customers want to buy again but cannot do so easily, improve the process.
If customers were disappointed with the first purchase, address the experience.
If customers are satisfied but have a genuine choice between similar suppliers, a relevant incentive may influence their next decision.
These situations require different investments.
Before spending money to encourage another purchase, investigate what prevented it.
Customer Intelligence Before Customer Incentives
Many businesses already possess useful information about why customers return or leave. The challenge is bringing that information together.
Purchase records tell you what happened. Customer conversations can help explain why.
Start by examining the customers who purchased once and those who returned. Compare their products, purchase intervals, service interactions, acquisition channels, and initial experiences.
Then investigate the differences.
A useful customer retention analysis should answer several questions:
- Which products have naturally high or low repeat-purchase rates?
- How long does a typical customer wait before buying again?
- Do customers who contact support return at different rates?
- Which complaints appear most often among customers who stop purchasing?
- Are frequent buyers responding to incentives or purchasing according to established habits?
- What do inactive customers say would have made them return?
Customer surveys, post-purchase calls, service records, and win-back research can provide answers that transaction data alone cannot.
For example, imagine a business discovers that first-time customers frequently contact support because they cannot understand product installation instructions.
Marketing could introduce a 10% second-purchase discount.
Or the company could improve onboarding, provide clearer instructions, and contact customers shortly after delivery.
Both approaches deserve evaluation. But only one directly addresses the problem customers have reported.
The right decision depends on evidence.
A 2021 study published in Marketing Science found that a non-tiered loyalty program increased estimated customer value by nearly 30% over five years in its studied setting. More than 80% of that estimated improvement came from reduced attrition rather than increased purchase frequency. PubsOnline
This illustrates why retention strategy should examine more than the number of transactions.
Keeping customers engaged, reducing preventable departures, and supporting successful product use may create value even when purchase frequency changes little.
For Idea Factor, this is where customer intelligence connects with customer operations. Research, customer feedback, service interactions, and CRM records should help businesses make coordinated decisions.
The objective is to identify the intervention most likely to change customer behaviour, rather than choosing a reward before understanding the problem.
How to Measure Whether a Loyalty Program Works
The financial case for a loyalty program begins before launch.
You need to know how customers behave without the proposed incentive.
Establish a baseline for purchase frequency, repeat purchase rate, retention, contribution margin, and customer lifetime value. Separate customers into meaningful segments.
Then test whether the program changes those outcomes.
A useful measurement framework includes six indicators.
| Metric | What it tells you |
|---|---|
| Repeat purchase rate | Whether more customers return within an appropriate purchasing window |
| Incremental purchase behaviour | Whether the program generates purchases beyond what would otherwise occur |
| Incremental contribution margin | Whether additional sales produce profit after relevant costs |
| Reward redemption rate | Whether customers actually receive and use meaningful rewards |
| Retention by segment | Which customers respond and which do not |
| Customer lifetime value | Whether the intervention improves long-term customer economics |
The most important word here is incremental.
A program can increase reported sales among members while producing little additional profit. Discounts may reduce the margin on purchases customers would have made at full price.
A simple example of incremental value
Imagine you have 1,000 eligible customers. You randomly assign 500 to receive a second-purchase incentive and 500 to a control group that does not receive it.
Hypothetical example, not research findings
Incentive group
30%
150 of 500 customers purchase again
Control group
20%
100 of 500 customers purchase again
Incremental purchases attributable to the test
50 additional purchases
A 10-percentage-point increase in repeat purchase rate.
Now examine profitability.
Suppose each purchase produces $40 in contribution margin before the incentive. The 50 additional purchases generate $2,000 in contribution.
But you also spent $10 on each of the 150 customers who purchased in the incentive group, including customers who would have returned anyway. That costs $1,500.
If administration and delivery cost another $300, the program produces only $200 in incremental contribution.
The extra purchases look promising. The financial return is much less impressive.
And that calculation still needs to consider whether customers simply purchased earlier than usual, whether future spending changed, and whether the test represents a sustainable operating model.
This is why a holdout group matters.
A before-and-after comparison cannot reliably separate program effects from seasonality, price changes, customer mix, and other business conditions.
Random assignment helps establish a more credible comparison. If randomization is impossible, carefully matched groups and historical purchasing data may help, but their limitations should remain explicit.
A 2025 EY loyalty market study also illustrates the measurement challenge. Among surveyed corporations, 36% identified return on investment as a key loyalty-program performance criterion, while 23% identified incremental member spending compared with nonmembers. Those are reported measurement priorities, not evidence of program effectiveness. EY
For marketing leaders, the decision is straightforward.
Do not approve a loyalty program based solely on projected enrollment or member spending. Require a credible method for measuring additional profit.
Five Questions to Ask Before Funding a Loyalty Program
These questions can help your leadership team decide whether to launch, redesign, or stop a program.
1. What customer behaviour are we trying to change?
Define the outcome precisely.
Are you trying to increase second purchases, shorten the time between orders, reduce cancellations, improve renewal rates, or encourage purchases across categories?
A useful answer identifies a customer segment, a specific behaviour, a measurement period, and a financial objective.
“Improve loyalty” is too broad to guide investment.
“Increase profitable second purchases among first-time customers within 90 days” gives your team something measurable, assuming that window fits the product’s purchasing cycle.
2. Who needs an incentive, and who would buy anyway?
Look at historical purchasing patterns before offering rewards.
Identify frequent buyers, occasional customers, new customers, and customers showing signs of disengagement.
Avoid assuming your highest-spending customers offer the greatest opportunity for incremental growth.
Some may deserve recognition and relationship benefits, but additional discounts may have little effect on their purchasing decisions.
Your goal is to identify customers whose behaviour can realistically change.
3. What is preventing customers from returning?
Investigate before proposing a solution.
Review complaints, customer interviews, abandoned reorders, satisfaction feedback, and service resolution data.
If the primary barrier is price or competitive choice, a targeted incentive may help.
If customers face delivery problems, poor support, or confusing processes, operational improvements may offer a better use of the budget.
A useful answer explains the barrier and provides evidence that it matters to the target segment.
4. Can we prove the program creates incremental value?
Define the control group, baseline, test period, and financial calculation before launch.
Include reward costs, discounts, administration, technology, customer support, and any other costs caused by the program.
Measure incremental contribution margin alongside repeat purchases and retention.
A useful answer explains what will be compared, how results will be evaluated, and what level of additional profit justifies the investment.
5. What would make us change or stop the program?
Every program needs a review date and performance thresholds.
Decide what happens if enrollment grows but repeat purchases do not. Decide what happens if customers redeem rewards without generating additional contribution margin.
Consider whether the program should be redesigned for a different segment, simplified, temporarily paused, or discontinued.
Without clear criteria, a loyalty program can become a permanent expense supported by increasingly familiar dashboards.
A Better Retention Strategy Starts With Understanding Your Customers
Customer retention rarely belongs to one department.
Marketing may design the incentive. Finance evaluates the economics. Operations influences fulfillment and reorder processes. Customer service handles the experiences that determine whether customers want to return.
If these teams operate independently, the business risks investing in rewards while ignoring the reasons customers leave.
A stronger approach connects customer behaviour with customer feedback and operational performance.
That means examining what customers purchase, listening to what they say, identifying recurring problems, and testing changes against measurable business outcomes.
This is also where Idea Factor’s work in customer research, customer intelligence, omnichannel support, and customer operations becomes relevant.
A customer service conversation can reveal a recurring problem that marketing data does not explain. A post-purchase follow-up can identify confusion before it becomes dissatisfaction. Consistent CRM records can help teams understand what happened across multiple interactions.
These capabilities support better retention decisions because they help businesses understand the customer experience beyond the transaction.
Sometimes the evidence will support a loyalty program.
Sometimes it will support a targeted offer, better onboarding, improved service, or a simpler purchasing process.
The important thing is to choose based on the problem and verify the result.
Reward the Behaviour You Want to Create
Loyalty programs can create meaningful business value. Research supports that conclusion.
But a program should earn its budget through measurable changes in customer behaviour and profitability.
Before approving another investment in points, tiers, or rewards, determine who is likely to return without incentives, who needs encouragement, and what prevents customers from purchasing again.
Then test whether your proposed solution improves the outcome.
The objective is to give customers better reasons to come back and ensure those reasons create value for the business.
If your team needs a clearer picture of why customers return or leave, Idea Factor can help connect customer intelligence, service experience, and operational insight to a more informed retention strategy.






Leave a Reply