What E-Commerce Brands Get Wrong About Their Loyalty Economics

Most e-commerce brands treat loyalty as a marketing problem. Run a points programme, send a birthday discount, add a VIP tier. But loyalty is an economics problem first, and when businesses skip that part, they build programmes that cost more than they return. The gap between a loyalty programme that looks good and one that actually improves margin is usually found in a handful of structural mistakes that are easy to miss until the numbers stop adding up.

Key Takeaways

• Loyalty programmes that ignore redemption costs and breakage rates often produce negative margin on repeat customers.
• Repeat purchase rate alone is not a reliable signal of loyalty; it can reflect habit, price, or lack of alternatives.
• The most effective loyalty economics are built around reducing acquisition cost, not increasing points earned.
• Brands in Australia, Singapore, Canada, and the US consistently underestimate the lifetime value ceiling created by a weak post-purchase experience.
• Loyalty strategy works best when it is aligned with the unit economics of each product category, not applied uniformly across the catalogue.

Why Do So Many Loyalty Programmes Fail on the Numbers?

A 2023 analysis by Bond Brand Loyalty found that around 57% of consumers belong to five or more loyalty programmes, but actively engage with fewer than half of them. Enrolment is not participation, and participation is not profitability.

The arithmetic breaks down in a specific sequence. A brand launches a points programme. Customers earn points. A portion of those customers redeem. The brand discounts the redemption against revenue it might have earned at full price. Then comes the accounting question most brands skip: what was the actual incremental margin on those repeat transactions, after the cost of the loyalty infrastructure, the fulfilment, the customer service, and the redemption discount?

In many cases, the answer is uncomfortable. The customer who redeems points regularly can actually generate less net margin than an occasional customer who pays full price and asks for nothing.

This is not an argument against loyalty programmes. It is an argument for understanding the economics before building one.

What Gets Measured Instead of What Matters

The metrics that dominate loyalty programme reporting tend to be enrolment count, repeat purchase rate, and average order frequency. These are descriptive numbers. They describe behaviour. They do not explain margin.

Consider a Shopify brand selling across three product categories: a high-margin consumable, a mid-margin seasonal product, and a low-margin commodity that drives volume. A blanket points programme rewards every dollar equally. But the economics of each category are completely different.

Rewarding the commodity purchase the same way as the consumable means subsidising the lowest-margin transaction in the catalogue. Over time, the loyalty programme trains customers to concentrate their purchases in the category that hurts the brand most.

The metric that should be tracked instead is contribution margin per loyalty customer cohort, segmented by product category and redemption behaviour. Almost no mid-market e-commerce brand builds that model before launching a programme.

The Breakage Trap

Breakage refers to loyalty points that are earned but never redeemed. Retailers and airlines have historically counted on breakage as a revenue offset. When customers earn points they do not use, the brand effectively collects revenue without paying the reward.

But breakage is not reliably predictable in e-commerce. And the brands that build their programme economics around an assumed breakage rate often find that rate collapses the moment a redemption campaign goes live, or a competitor launches something more compelling and customers suddenly decide to cash out.

Airlines can model breakage because seat inventory is finite and expiry rules are strict. A DTC brand with rolling point expiries and broad redemption options has far less control over when and how liabilities are triggered.

Building loyalty economics around projected breakage is a gamble dressed up as a strategy.

Repeat Purchase Rate Is Not the Same as Loyalty

This distinction matters more than most brands acknowledge. A customer who buys from you every three months might be loyal. Or they might be buying because you are the cheapest option available, because your checkout is faster than your competitors, or because your category has no clear alternatives in their market.

When a genuine competitor enters, or when your price parity slips, those customers leave immediately. They were never loyal. They were just satisficed.

True loyalty shows up in a specific pattern: customers who return despite a price increase, who recommend the brand without incentive, and who resist switching even when a lower-cost option exists. Research by Bain and Company has found that increasing customer retention rates by 5% can increase profits anywhere from 25% to 95%, but that range reflects the enormous difference between habit-driven retention and genuine preference-driven loyalty.

If you want to understand where your brand actually sits on that spectrum, assessing your brand health across dimensions like advocacy, perceived value, and emotional connection is a useful starting point. You can get a baseline quickly using the free brand health score tool from Lenka Studio, which surfaces the gaps that repeat purchase rate alone will never show.

What Does Good Loyalty Economics Actually Look Like?

The brands that build sustainable loyalty programmes tend to share a few structural choices.

They focus on reducing acquisition cost, not inflating retention metrics.

A loyal customer who refers two new customers a year is worth dramatically more than a loyal customer who simply buys again. The referral collapses your customer acquisition cost on that new buyer. Programmes that reward referral behaviour explicitly, rather than just purchase frequency, create compounding returns instead of linear ones.

They design for high-margin behaviour, not all behaviour.

The best loyalty programmes in e-commerce are category-aware. They reward the actions that correspond to the highest-margin outcomes: subscribing to a replenishment plan, purchasing in bundles, reviewing a premium product, or upgrading to a higher-value tier. They do not reward every transaction equally.

They treat post-purchase experience as the loyalty mechanism, not the points system.

Research from PwC consistently shows that around 32% of customers stop doing business with a brand they love after a single bad experience. A points programme cannot compensate for a poor unboxing, a slow return, or an unhelpful support interaction. Brands that invest in post-purchase experience before they invest in loyalty infrastructure build a more durable retention advantage at lower cost.

They cap liability exposure before launch.

Structurally sound programmes define the maximum redemption liability they can absorb per quarter, and they build expiry rules, earning caps, and redemption thresholds accordingly. This is not about being stingy. It is about not creating an open-ended financial obligation that scales in an uncontrolled way as the programme grows.

The Lifetime Value Ceiling Most Brands Don't See Coming

Customer lifetime value is the metric loyalty programmes are supposed to increase. But there is a ceiling on LTV that most brands do not account for in their modelling.

That ceiling is set by the category's natural repurchase cycle, the realistic number of product lines a single customer will buy from you, and the gross margin per order. If a brand sells a consumable that a customer buys every eight weeks at $60 average order value with a 35% gross margin, the annual gross profit per customer is around $137. A loyalty programme that costs 6% of revenue to run reduces that to around $104. The room to grow LTV is genuinely narrow.

Brands that understand this ceiling early make different decisions. They consider whether the category can expand (are there adjacent products that would increase basket size?), whether the customer base can be segmented by higher-value cohorts, and whether the loyalty investment is better directed at reducing churn rather than increasing frequency.

For a Singapore-based beauty brand or a Canadian supplement company, the ceiling calculation looks quite similar. The ceiling is a product-category constraint, not a geography one.

Where Agencies and In-House Teams Often Diverge on This

Building the right loyalty economics model requires skills that sit at the intersection of commercial analysis, data engineering, and retention strategy. Most in-house e-commerce teams are strong on one of those. Rarely all three at once.

A team that is excellent at CRM and email tends to focus on the engagement metrics. A team that is commercially sharp tends to focus on the margin story but may lack the technical infrastructure to track it at the cohort level. A data team can build the model but may not know which commercial questions to ask first.

At Lenka Studio, the loyalty work we do for e-commerce clients typically starts with the economics model before any programme mechanics are designed. Getting the unit economics right before the first point is issued saves significant rework later, and it prevents the very common situation where a brand is twelve months into a programme and cannot tell you whether it is working in any meaningful financial sense.

What the Numbers Tend to Reveal in Practice

Across mid-market e-commerce brands, a few patterns show up repeatedly when the loyalty economics are properly modelled.

• Around 60 to 70% of enrolled loyalty members never redeem, which inflates retention metrics without reflecting genuine engagement.
• The top 10% of loyalty customers often generate 30 to 40% of programme redemption cost, meaning a small cohort drives a disproportionate liability.
• Brands that restructure programmes around referral and review incentives rather than purchase points typically see 15 to 25% improvement in new customer acquisition cost within 12 months.
• Post-purchase NPS scores below 7 reliably predict churn within two purchase cycles, regardless of points balance held.

These are not industry-wide published statistics. They are patterns from commercial modelling work. The specific numbers will differ by category, AOV, and market. But the direction is consistent.

Frequently Asked Questions

If you are building a loyalty programme or rethinking one that is not performing, talk to the team at Lenka Studio. We work with e-commerce brands across Australia, Singapore, Canada, and the US to build the commercial foundation before the customer-facing strategy, so the programme works on both the balance sheet and the brand.