Loyalty programmes: design and economics

What is a loyalty programme supposed to change?

A loyalty programme should make an already useful relationship more valuable for both sides. It might encourage a second purchase, help customers discover a relevant service or recognise dependable repeat business.

Points are only one delivery mechanism. A programme can offer repairs, priority assistance, convenient replenishment, access to appointments or a clearly priced membership. The benefit should solve a recognisable problem. A reward that requires customers to study a complicated conversion table is a weak starting point.

Separate loyalty from lock-in. A customer who returns because a service is dependable has a different relationship from one who returns only to avoid losing a balance. Both behaviours may produce transactions, but they carry different expectations and different risks when the rules change.

This article focuses on programme design. The broader client retention framework covers the service, communication and delivery work that must exist before a reward mechanism can help.

Why do loyalty programme economics need scrutiny?

Rewards create costs immediately, while the behaviour they are meant to change may already have happened without them. The programme therefore needs an incremental business case, not a chart showing that members spend more.

The US Small Business Administration's break-even guide, checked in October 2026, distinguishes fixed costs from the contribution available on each sale. That distinction is useful here: extra revenue can look attractive while reward fulfilment consumes the contribution needed to pay for the programme.

The IFRS Foundation's IFRS 15 overview also ties revenue recognition to contractual promises. Finance should assess how a proposed benefit affects the applicable accounting treatment. Marketing should not assume that cash collected equals revenue available to spend, or that unredeemed rewards have no consequences.

These are planning principles, not a prescription for a particular accounting policy. An Indian business needs its finance adviser to apply the standards relevant to that business and its actual programme terms.

How do you design and test a loyalty programme?

Build the programme around one behaviour, one understandable benefit and one measurement plan. Add complexity only when customer evidence justifies it.

1. Define the behaviour and the customer problem

Choose a specific bottleneck. A refill business may need customers to remember when stock runs low. A service business may need a smoother renewal decision. A fashion retailer may need customers to find suitable products after the first purchase.

Do not treat all three problems as a discount problem. Reminders, advice and easier reordering may do more useful work than subsidising the next transaction. Interview customers who returned and those who did not before deciding that a reward is the missing ingredient.

2. Write the exchange in ordinary language

Explain what customers do, what they receive, when it becomes available and what restrictions apply. Give staff examples of a successful redemption and an unsuccessful one. If the explanation changes across the website, till and support desk, the programme is not ready.

Model cancellations and refunds as well as successful purchases. Decide what happens when a qualifying order is returned after the reward has been used. Give customers a visible balance and a way to correct errors without a lengthy argument.

3. Calculate contribution after the benefit

Consider an illustrative order with net revenue of ₹1,000, variable product and delivery costs of ₹650 and a reward cost of ₹50. These are assumptions, not industry benchmarks. Contribution after the reward is ₹300 before programme overhead and acquisition costs.

If that order would have happened anyway, the reward reduced contribution by ₹50. If the programme caused a genuinely additional order, it may have added ₹300. The same transaction therefore has a very different economic interpretation depending on the counterfactual.

Include technology fees, fraud, customer support, partner reimbursement and the cost of communicating benefits. Show a downside case with higher redemption and lower additional purchasing. Do not make profitability depend on customers misunderstanding or forgetting their entitlement.

4. Create a credible comparison

Where practical, randomly offer a pilot to some eligible customers and compare them with customers receiving the normal service. Set the purchase window before starting. Compare contribution per eligible customer, not merely average spending among those who joined.

If randomisation is impractical, use a carefully matched comparison and state the limitation. Customers who volunteer for a membership may already be more engaged. A before-and-after increase can also reflect seasonality, better availability or a simultaneous promotion.

5. Review the experience alongside the margin

Track reward comprehension, successful redemption and support complaints. A profitable pilot with persistent confusion needs repair. A popular pilot with poor economics needs a different benefit or a clearer commercial purpose.

The CCPA's December 2023 announcement on dark patterns identifies deceptive practices including false urgency and drip pricing. Have counsel review the actual journey where relevant. The practical design standard is straightforward: customers should understand the exchange before committing.

Where do loyalty programmes fit in India?

They fit best where customers have a natural reason to return and the business can deliver the benefit consistently. A neighbourhood business with reliable records may operate a simpler programme than a national retailer, without producing a worse customer experience.

Design for assisted use as well as self-service. A customer buying through a shop, phone call or messaging conversation should not lose a valid reward because the system recognises only app purchases. Resolve identity carefully rather than forcing people to disclose unnecessary information.

Language also matters. Explain expiry and eligibility in the language used to sell the programme. Do not make the benefit prominent in a regional-language creative while leaving the restriction understandable only to an English-speaking customer.

A low-frequency purchase may not support a traditional points model. Installation support, maintenance information or relevant referrals may create more value. In professional and regulated services, assess whether incentives could distort an important decision before offering them.

When should a programme launch or change?

Launch after ordinary purchasing, fulfilment and complaint handling work reliably. A reward cannot compensate indefinitely for missing stock or unresolved refunds.

In week one, map the purchase cycle and interview customers. Next, cost a simple benefit and test staff understanding. Run the pilot for a period that includes a realistic opportunity to repeat, rather than choosing a convenient reporting month.

Review changes before communicating them. Customers may have planned around existing benefits. Explain transitional arrangements, give a clear contact route and retain a record of the terms that applied when someone joined. Ask finance and legal reviewers to assess material changes before release.

Who owns the promise after launch?

One programme owner should coordinate marketing, finance, technology and operations. Marketing controls the proposition; it cannot independently promise benefits that other teams must fund or fulfil.

Finance should approve the contribution model and review actual costs. Operations should test redemption at the point of service. Technology should maintain the reward ledger and exception handling. Support should report recurring misunderstandings, not simply close individual tickets.

Partners need explicit responsibilities too. If another business supplies the reward, establish what happens when it is unavailable. A customer experiences the brand's promise, regardless of which supplier failed behind the scenes.

My view

I would prefer a modest benefit customers can reliably use over a dramatic points balance with little practical value. The trade-off is that simpler programmes can look less exciting in a launch presentation. They are often easier to explain, operate and evaluate.

I would also keep a route for non-members to receive competent service. Membership should add value without making the ordinary customer experience deliberately frustrating.

A reward earns its place when it improves the relationship and survives the margin calculation.

Start with the second purchase

Identify why a satisfied first-time customer might not return. Write one proposed benefit, its fully loaded cost and the behaviour it should change. If the team cannot explain how it will distinguish additional business from subsidised existing business, resolve that before buying a loyalty platform.