Loyalty programmes: most fail, and the ones that work are simple

Most loyalty programmes fail in the same way: they are launched, they are complicated, staff stop mentioning them after three weeks, and a year later they are a line of cost with a handful of active members.

The test before you design anything

Can a cashier explain it in one sentence, to a customer who is in a hurry?

"Buy nine coffees, get the tenth free" passes. "Earn points on every purchase which convert to tier credits redeemable against selected items" does not.

If it fails this test, it will not be used, regardless of how good it looks in a specification.

The three models

Stamps — buy N, get one free

Best for: cafés, bakeries, fast-moving repeat purchases.

Why it works: instantly understandable, visible progress, feels like a gift rather than a discount.

Cost: one item in N. At a 65% margin on coffee and nine purchases before the reward, the cost is about 7% of revenue from members. Work out your own.

Points — spend, accumulate, redeem

Best for: higher-value retail with varied basket sizes.

Why it works: it scales with spend, so a large customer earns more.

The trap: points people never redeem are a liability and a disappointment. Keep the rate simple enough that a customer can do the arithmetic themselves.

Tiers — spend more, get better treatment

Best for: categories where service is part of the product — jewellery, electronics, high-end retail.

Why it works: the reward is status and treatment rather than discount, which costs less margin.

Requires: you actually deliver the different treatment. A tier that promises priority service and does not deliver it is worse than no tier.

Wameed supports tiers, badges and challenges as well as the simpler models; see thecustomer features.

What makes them work

Enrolment takes five seconds. A phone number at the counter. Not a form, not an app download. Seedigital receipts — sending the receipt and enrolling can be the same question.

Progress is visible. The customer should be told where they are at each purchase: "that's your seventh."

The reward is worth having but not extravagant. A free coffee after nine is worth having. 2% off after twenty is not worth anyone's attention.

Staff mention it. The whole programme runs on the cashier asking. If they stop asking, it stops existing — which makes this a staffing and coaching issue more than a software one.

It expires slowly or not at all. Aggressive expiry generates more resentment than it saves in liability.

The consent part

A phone number is personal data under the Personal Data Protection Law. Enrol with a clear statement of purpose, ask separately before sending marketing, and be able to delete on request. Seecustomer data and PDPL.

Measuring whether it works

Three numbers, and only three:

1. Enrolment rate — proportion of transactions from enrolled members. Below 20% after three months means staff are not asking.
2. Repeat rate: members versus non-members. This is the whole point. If members do not visit more often than non-members, the programme is a discount, not a loyalty programme.
3. Cost of rewards as a percentage of member revenue. Compare it to the uplift in number 2.

If number 2 shows no difference after six months, stop the programme. It is costing margin and changing nothing, and continuing it out of sunk cost is a common and expensive mistake.

The simplest thing that works

For most small Saudi shops, this is enough:

• Capture the phone number when sending the digital receipt
• A stamp card in the POS, tied to the number, with no physical card to lose
• One clear reward
• A staff prompt at the counter
• A monthly look at the three numbers above

No app, no tiers, no points arithmetic. It takes an afternoon to configure and it outperforms most elaborate programmes, because it is the one staff will actually use.