How to Build a Customer Loyalty Programme
Loyalty programmes fail for predictable reasons. The reward is too distant to motivate anyone, customers cannot see their balance, the maths quietly eats the margin, or the programme rewards people who would have bought anyway.
This guide treats loyalty as a system with an economic model, not a marketing gesture. It covers the mechanics, the arithmetic, the technical build and the Nigerian realities that decide whether customers actually engage.
When a loyalty programme is worth building
A loyalty programme rewards repeat purchasing to increase how often customers buy, how much they spend and how long they stay. It only pays back when repeat purchase is realistic in your category.
Run this test before spending anything:
| Question | Loyalty likely to work | Loyalty likely to waste money |
|---|---|---|
| How often does a typical customer buy? | Monthly or more often | Once every few years |
| Is your category competitive on convenience? | Yes, switching is easy | No, you are the only supplier |
| Can you identify customers across purchases? | Yes, by phone number or account | No, mostly anonymous walk-ins |
| Is your margin able to fund a reward? | Yes, at least a few percent to give | Very thin, commodity pricing |
| Do you already have repeat customers? | Yes, some | No, the product itself is the problem |
Categories where loyalty programmes usually justify the build: pharmacy, groceries, beauty and skincare, food and drink, laundry, fuel and lubricants retail, printing, salon and barber services, pet supplies, baby products.
Categories where it rarely does: furniture, wedding services, property, heavy machinery, and any business where the same customer buys once every few years. There, referrals are the better investment.
If your problem is that first-time buyers never return because delivery was late or the product disappointed, fix that first. A loyalty programme cannot buy back a bad first experience.
Choose one mechanic, not four
| Mechanic | How it works | Strengths | Weaknesses | Best fit |
|---|---|---|---|---|
| Points | Earn points per naira spent, redeem for discounts or items | Flexible, familiar | Feels abstract; needs visible balances | Retail with frequent purchases |
| Cashback wallet | A percentage returns to a store wallet usable next time | Concrete and easy to explain | Feels like money owed; needs clear rules | Online stores, groceries |
| Tiers | Spend thresholds open better benefits | Rewards your best customers | Only motivates those near a threshold | Beauty, fashion, services |
| Stamp or punch card | Buy a set number, get one free | Very simple, works offline | Easy to game without identity checks | Food, laundry, salons, coffee |
| Paid membership | Customer pays for delivery or discount benefits | Predictable revenue, strong retention | Only works with genuine frequency | Groceries, pharmacy, subscription-style buying |
| Surprise rewards | Unannounced credits or gifts for good customers | Cheap, generates goodwill | Not predictable enough to change behaviour | Any business, as a supplement |
Choose one primary mechanic. Programmes that mix points, tiers, stamps and coupons confuse customers and complicate the build. Add a second mechanic only after the first is understood and used.
A simple decision rule: if customers buy small amounts frequently, use stamps or cashback. If basket sizes vary widely, use points. If you want to protect high-value customers from competitors, use tiers.
The economics: earn rate, burn rate and margin
This is where most programmes go wrong quietly.
Define three numbers before anything is built:
- Earn rate: what the customer receives per naira spent, expressed as a percentage of spend.
- Burn rate or redemption value: what a point is worth when redeemed.
- Expected redemption: the share of issued rewards you expect to be claimed. Unclaimed rewards are called breakage.
Illustrative arithmetic (your numbers will differ):
| Item | Illustrative figure |
|---|---|
| Average order value | ₦15,000 |
| Gross margin | 30%, or ₦4,500 per order |
| Earn rate | 3% of spend, so ₦450 per order in rewards |
| Expected redemption | 60% of issued value |
| Expected real cost per order | ₦270 |
| Cost as a share of margin | 6% of the ₦4,500 margin |
The programme then needs to lift behaviour by more than that cost. If a member buys five times a year instead of four, the extra order's margin easily covers the reward cost across all five orders. Model that before launch, and re-check it after a quarter using your own data.
Four financial rules:
- Fund the reward from margin, not revenue. A 5% reward on a 10% margin is half your profit.
- Cap the maximum discount per transaction, for example no more than 20% of any single order paid with points, so a large redemption does not produce a loss-making sale.
- Track the liability. Issued-but-unredeemed points are a future cost. Report the balance monthly, and discuss the accounting treatment with your accountant.
- Set an expiry, commonly 6–12 months of inactivity, communicated clearly. Expiry limits liability and encourages return visits, but aggressive expiry breeds resentment, so warn customers before points lapse.
Design rules that decide whether it works
- The first reward must be reachable quickly. If it takes ten purchases to feel anything, most customers never arrive. Aim for a visible benefit within two to four purchases.
- Make the value obvious in naira. "You have 2,400 points" means nothing. "You have ₦2,400 to spend" does.
- One enrolment step. A phone number at checkout is enough. Do not require an app download or a long form.
- Reward more than spending. Points for a review, a referral, a completed profile or choosing self-pickup shape useful behaviour and cost less than discounts.
- Make redemption easy at the moment of purchase, not through a separate claim process.
- Keep the rules on one page in plain language and link it everywhere.
- Give staff a reason to promote it. In-store, the cashier decides whether the programme lives. Track enrolments per cashier and recognise the best.
How to build it: the technical parts
Five components, whatever route you take:
1. Customer identity. One record per customer, keyed on the phone number, with merge handling for duplicates. In Nigeria the phone number is the practical identifier because it works online, in store and on WhatsApp. Normalise formats (+234 and 0 prefixes) or you will create duplicate accounts and double-issue points.
2. Points ledger. Never store a single balance field. Store every transaction — earned, redeemed, expired, adjusted — with the reference order, timestamp and reason. The balance is the sum. This is what lets you resolve a dispute and audit the liability.
3. Rules engine. Earn rules (rate, excluded categories, bonus multipliers), redemption rules (minimum balance, maximum share of an order, eligible products), and expiry rules. Keep these configurable in an admin screen; you will change them, and you should not need a developer each time.
4. Touchpoint integrations. Website checkout, point of sale for in-store purchases, WhatsApp or Instagram orders entered by staff, and any app. Every channel must read and write to the same ledger, or customers will report balances that do not match.
5. Communication. Balance on receipts and order confirmations, a monthly WhatsApp or SMS summary, reminders before expiry, and a message when a reward becomes available. Honour opt-outs and keep the privacy notice current under the Nigeria Data Protection Act 2023.
Where customers check their balance
Visibility is the single biggest driver of engagement, and it is where Nigerian design differs from imported templates.
- On the receipt and the order confirmation. Cheapest and most reliable.
- By WhatsApp. A "check balance" reply handled by a simple automated flow suits customers who will never open a portal.
- On the account page of your website, for online buyers.
- By USSD or SMS keyword, for customers with limited data.
- In the app, if you already have one with real usage. Do not build an app for the loyalty programme alone.
Every balance display should show what the customer can get right now, not only the number: "₦2,400 available — enough for a free delivery or ₦2,400 off your next order".
Fraud and abuse controls
- Require the phone number at the point of sale, not after, so staff cannot attach strangers' purchases to their own numbers.
- Block self-referral and same-device abuse where a referral element exists.
- Cap points earned per day per customer.
- Require the points to be earned before they can be redeemed, with a short clearing delay for online orders so refunds can reverse them.
- Reverse points automatically on refunds and returns.
- Review the top 20 earning accounts monthly. Unusual patterns are usually staff-related.
How much does a loyalty programme cost to build?
| Route | What you get | Indicative cost |
|---|---|---|
| Platform plugin or app on your existing store | Basic points or cashback, account page display, email notifications | ₦150,000–₦800,000 plus monthly fees |
| Configured loyalty tool integrated with store and POS | Rules engine, in-store and online earning, reporting | ₦800,000–₦2,500,000 plus subscriptions |
| Custom loyalty engine | Your own ledger, rules, tiers, WhatsApp balance checks, fraud controls, full reporting | ₦1,500,000–₦6,000,000 |
| WhatsApp balance and notification layer | Automated balance checks and campaign messages | ₦300,000–₦1,200,000 plus per-message costs |
Indicative 2026 ranges; actual quotes vary with scope, vendor and exchange rate. Recurring costs include messaging fees, maintenance at ₦20,000–₦150,000 per month and, of course, the reward cost itself, which is the largest ongoing line and should sit in the marketing budget.
What changes for Nigerian customers
Price sensitivity favours immediate, concrete value. A naira-denominated wallet balance or a free delivery beats abstract points. Where a discount and a reward compete, many customers prefer the certain discount, so the programme must feel like money.
Phone numbers are the identity layer. Email addresses are unreliable for many segments; phone numbers are used for payments, WhatsApp and delivery.
WhatsApp is the loyalty channel. Balance checks, reward alerts and expiry reminders land where customers already are. Business-initiated messaging at scale runs through the WhatsApp Business Platform with approved templates and per-message costs.
Data costs limit app-based programmes. A programme that requires an app install excludes customers who will not spend data on it. Web plus WhatsApp reaches more people.
In-store enrolment needs staff incentives. In supermarkets, pharmacies and salons, the cashier's habit determines enrolment rates.
Airtime and data make popular rewards because their value is unambiguous. If you offer them, verify the reliability and cost of any airtime distribution provider before promising instant delivery.
Trust must be earned early. Customers who have seen promotions evaporate will test your programme with a small redemption. Make the first redemption effortless.
Example (hypothetical): a pharmacy chain in Port Harcourt
Example (hypothetical): a three-branch pharmacy chain with an online store wants repeat customers for chronic medication and everyday health items.
Programme design:
- Mechanic: cashback wallet, 3% of spend, excluding items with regulated or very thin margins.
- Identity: phone number captured at the till and at online checkout.
- Redemption: usable from ₦1,000, capped at 25% of any single purchase.
- Expiry: 12 months from the last transaction, with a WhatsApp reminder 30 days before.
- Bonus earning: double cashback on scheduled refill purchases, encouraging adherence and predictable demand.
- Visibility: balance printed on every receipt, plus a WhatsApp balance check.
- Build: loyalty ledger integrated with the POS and the online store, indicative ₦2,400,000, about ten weeks.
The design decision that mattered most was excluding regulated and thin-margin items from earning. Without that exclusion, the programme's cost would have concentrated on exactly the products that could not fund it.
What to measure
Review monthly, and compare members against non-members rather than looking at member numbers alone:
- Enrolment rate: share of transactions attached to an identified customer
- Repeat purchase rate for members versus non-members
- Purchase frequency: average orders per member per quarter
- Average order value for members versus non-members
- Redemption rate: share of issued value redeemed
- Outstanding liability: issued value not yet redeemed or expired
- Programme cost as a percentage of gross margin
- Reactivation: dormant members who returned after a reminder
If members and non-members behave identically after two quarters, the programme is a discount, not a loyalty mechanism, and the design needs changing.
Mistakes to avoid
- Rewards too far away. A first reward at ten purchases guarantees disengagement.
- Hidden balances. If customers cannot see their points without asking, they will not change behaviour.
- Funding rewards from revenue rather than margin. It works until you look at the accounts.
- No expiry and no liability tracking. The cost accumulates silently for years.
- Points that survive refunds. Reverse them automatically or you create an arbitrage.
- Launching without staff training. In-store enrolment collapses if the cashier finds it slow.
- Complex rules. If the programme needs a paragraph to explain, simplify it.
- Letting the programme substitute for fixing delivery, stock or product problems. Loyalty amplifies a good experience; it cannot replace one.
Launch checklist
- Repeat-purchase test passed for your category
- One mechanic chosen and written on a single page in plain language
- Earn rate, redemption value and cap modelled against gross margin
- Points ledger with full transaction history, not a balance field
- Phone number normalisation and duplicate merging tested
- Earning and redemption working on website, POS and chat-order channels
- Balance visible on receipts, order confirmations and WhatsApp
- Refund reversal and expiry rules implemented
- Fraud caps and staff-abuse monitoring in place
- Privacy notice updated and opt-outs honoured
- Staff trained and enrolment tracked per branch or cashier
- Monthly reporting pack defined before launch
Conclusion
A loyalty programme is an economic mechanism with a user interface. Confirm that your customers buy often enough to justify it, pick a single mechanic they can explain to a friend, set earn and redemption rates your margin can carry, build a proper ledger keyed on the phone number, and make the balance impossible to miss. Measure members against non-members, and change the design if the difference does not appear.
If you want a loyalty programme that works across your website, your shop counter and WhatsApp without the balances disagreeing, Linestech builds loyalty and customer data systems for Nigerian businesses.
Frequently asked questions
Points or cashback: which works better in Nigeria?
Cashback expressed in naira is usually easier for customers to value, because the benefit is unambiguous. Points suit businesses with wide basket variation or where rewards include non-cash items. Whichever you choose, display the balance in naira terms so nobody has to do mental arithmetic.
Do I need an app for a loyalty programme?
No. Phone-number identity plus a web account page and WhatsApp balance checks reach far more Nigerian customers than an app, and cost much less to build. Add app support only if you already have an app with regular usage.
How much should I give away?
Most programmes sit between 1% and 5% of spend, funded from gross margin and adjusted for expected redemption. Start conservatively; increasing a reward later is easy, while reducing one damages trust.
Should points expire?
Usually yes, commonly after 6 to 12 months of inactivity, to limit liability and encourage return visits. Communicate expiry clearly at enrolment, remind customers before points lapse, and never expire points without notice.
How do I run a loyalty programme for customers who buy in person with cash?
Capture the phone number at the till and record the sale against it in the same system as online orders. Print the balance on the receipt. Cash customers are often your most frequent buyers, and excluding them defeats the purpose.
Can a small business with 200 customers run one?
Yes, and simply. A phone-number list, a stamp or cashback rule, balances recorded in your sales system and a monthly WhatsApp message can work well. Build custom software when manual tracking starts producing errors or takes real staff time.
How do I stop staff from abusing the programme?
Require the customer's number before the sale is completed, cap daily earning per number, block staff numbers from earning on customer transactions, and review the highest-earning accounts monthly. Most abuse shows up as one number attached to an implausible number of transactions.
Sources and further reading
Figures, platform rules and regulations change. These are the primary references behind this article and the places to check before you act on it.
