How to Build a Referral Programme

Referrals already happen in most Nigerian businesses. Someone posts your product in a family WhatsApp group and three people order. The difference a programme makes is that you can see it, reward it and repeat it deliberately, instead of hoping.

This is a build guide, not an encouragement to "ask for referrals". It covers the reward maths, the attribution mechanics, the fraud controls and the sequence that gets a programme live.

What a referral programme is and when it beats advertising

A referral programme is a structured system that gives existing customers a reason and a mechanism to bring new customers, tracks who brought whom, and pays a reward when the new customer meets a defined condition.

It tends to outperform paid advertising when:

• Your product requires trust that an advert cannot establish — food, health, childcare, finance, home services.
• Customers naturally discuss the category with peers.
• Your margins per customer are healthy enough to share.
• Your advertising costs per acquired customer are rising.

It works poorly when the product itself disappoints, when purchases are rare and private, or when your customers have no obvious peer group with the same need.

Referral differs from three neighbours, and mixing them up creates confusion:

Start with referral. Affiliates and agents need contracts, tax treatment and closer management.

Choose the reward structure

For most Nigerian consumer businesses, two-sided works best, because it gives the referrer something to offer rather than something to ask for. "Use my code and get ₦1,500 off your first order" is a gift; "refer me and I get ₦1,500" is a favour.

Choose reward types carefully:

• Store credit is cheapest because it returns as a future purchase, and it costs you margin rather than cash.
• Cash or bank transfer is the strongest motivator and the most abused. Use it for high-value products only, with verification.
• Airtime or data is popular and easily understood; confirm the reliability and cost of any distribution provider before promising instant delivery.
• Free delivery is powerful in categories where delivery fees affect the decision.

Reward economics: what you can afford to pay

Work backwards from what a customer is worth to you.

Illustrative arithmetic (substitute your own figures):

Three rules:

1. Compare against your real acquisition cost, not against zero. Referrals are cheap, not free.
2. Pay on qualification, not on sign-up. Rewarding registrations attracts people collecting rewards, not customers.
3. Cap the total exposure per referrer per month until you understand the patterns.

If you do not know your acquisition cost, use gross margin per first order as a ceiling: the combined reward should not exceed the margin on the first purchase it generates, unless you have evidence that referred customers repeat.

Attribution: how referrals get tracked

Attribution decides whether the programme is measurable or merely hopeful.

The practical Nigerian combination is code plus link plus a "who referred you?" field at checkout, with a short manual claim window as a safety net. Assume that a meaningful share of referrals will arrive by word of mouth and never touch the link.

Design details that matter:

• Make codes short, readable and memorable, ideally based on the customer's name.
• Ensure links carry the code through to checkout even if the customer browses away and returns later.
• Use the first referrer recorded, not the last, and say so in the rules.
• Log every attribution event so disputes can be answered with facts.

Define what counts as a qualified referral

Write this down before launch, because it is the source of most arguments.

A workable definition: a referral qualifies when a new customer, not previously in your database by phone number or email, places a paid order above a minimum value, and the return or cancellation window has passed.

Set each element explicitly:

• New customer test: phone number and email not in the database, and delivery address not matching an existing account. Address matching catches household abuse.
• Minimum order value: high enough that the reward is not larger than the margin.
• Holding period: commonly 7–14 days after delivery, so refunds do not leave you paying for a reversed sale.
• Exclusions: staff, existing customers returning under a new number, bulk or wholesale orders.
• Expiry: codes that stop working after a promotional period, if you want control.

When and where to ask

Timing changes results more than reward size.

The best moments:

1. Immediately after a delivered order is confirmed, when satisfaction is highest.
2. After a positive review or rating.
3. On the second or third purchase, when the customer has demonstrated they like you.
4. When a reward is redeemed, because they are already thinking about value.

The best places for a Nigerian consumer business:

• A WhatsApp message with the code, a short line the customer can forward, and a link that opens your store.
• The order confirmation and delivery notification messages.
• The account page and the post-checkout thank-you page.
• The packaging insert, with a code and a QR code, for physical products.
• In-store at the counter, if you have a shop.

Give people the exact words to forward. Most customers will not compose a recommendation; they will forward one.

The technical build

Five components:

1. Referral record. Referrer identity, code, channel, referred contact where known, and a state: shared, clicked, signed up, ordered, qualified, rewarded, rejected. Never collapse this to a single flag.

2. Code and link generation. One persistent code per customer, generated at first purchase or at account creation, plus a shareable link.

3. Matching engine. Applies the code at checkout, checks the new-customer test, records the attribution and starts the holding period.

4. Reward issuance. Credits a store wallet, issues a coupon, or queues a cash payout after qualification. Store credit should post to the same ledger as any loyalty programme so balances do not conflict.

5. Dashboards. A simple customer view ("you have referred 4 people, 3 qualified, ₦4,500 earned") and an internal view showing programme cost, qualified referrals and suspicious patterns.

Keep the qualification rules configurable in admin. You will adjust the minimum order value and holding period within the first quarter.

Fraud and abuse controls

Referral programmes attract gaming wherever cash is involved.

• Block self-referral by phone number, email, device and delivery address.
• Require the referred order to be paid, not pay-on-delivery, before a cash reward, or extend the holding period until remittance is confirmed.
• Cap rewards per referrer per month and per day.
• Flag clusters: many referrals to the same address, the same device fingerprint, or numbers registered minutes apart.
• Reverse rewards automatically when the referred order is refunded or returned.
• Reserve manual review for cash payouts above a threshold.
• Exclude staff and their immediate contacts explicitly in the rules.

State in the programme terms that you may withhold rewards for abuse, and define an appeal route so honest customers are not lost to a false positive.

How much does a referral programme cost to build?

Indicative 2026 ranges; actual quotes vary with scope, vendor and exchange rate. Running costs are the rewards themselves, messaging fees and maintenance at ₦20,000–₦150,000 per month shared with your store.

What changes for Nigerian referral programmes

WhatsApp is the distribution channel. Forwarded messages in family, estate, church, office and market association groups move faster than any social advert. Design the share message for forwarding: one line of context, the offer, the code and a short link.

Group sharing means the referrer may not know the referee. Attribution by code survives this; attribution by personal invitation does not.

Cash rewards carry more weight but more risk. Bank transfer payouts are easy through Nigerian payment providers, which is exactly why they attract abuse. Store credit is safer and often sufficient.

Trust is the currency. A referral from a known person overcomes the hesitation many Nigerian shoppers feel about paying an unfamiliar online seller. That is why referred customers often convert more readily than advertising traffic.

Pay-on-delivery complicates qualification. An order placed is not an order paid. Tie the reward to confirmed remittance rather than to dispatch.

Data protection still applies. If customers submit their friends' phone numbers, you are collecting third-party personal data. Under the Nigeria Data Protection Act 2023, be transparent about what you do with it, avoid unsolicited marketing to contacts who never engaged, and give people a way to be removed.

Example (hypothetical): a meal-prep business in Lagos

Example (hypothetical): a Lagos meal-prep service delivering weekly food packs to about 300 subscribers wants growth without raising advertising spend.

Programme design:

• Structure: two-sided. The new customer gets ₦2,000 off their first weekly pack; the referrer gets ₦2,000 in store credit after the referred customer's second delivery.
• Why the second delivery: first-week churn is high in meal services, so qualification at delivery two rewards genuine customers rather than trial-takers.
• Attribution: a personal code such as ADA2K, a link, and a "who told you about us?" field at checkout with a phone number.
• Ask timing: a WhatsApp message the evening after each successful delivery, with a forwardable line and a picture of that week's menu.
• Controls: one reward per new household address, credit only for the referrer, monthly cap of five rewarded referrals.
• Build: referral module on the existing store, indicative ₦1,300,000, six weeks.

The decision worth copying is the qualification point. Moving the reward from sign-up to the second delivery removed most of the abuse and aligned the programme with the behaviour that actually generates profit.

What to measure

• Share rate: percentage of customers who share a code or link
• Referral conversion: qualified referrals divided by referred visitors
• Cost per acquired customer through referral, versus your other channels
• Repeat rate and average order value of referred customers versus others
• Programme cost as a share of gross margin generated by referred customers
• Rejection rate and reasons, which reveal fraud pressure and rule problems
• Time from share to qualification, which tells you whether the holding period is right

Review after one full purchase cycle. If referred customers repeat at least as well as other customers, increase the reward; if they repeat worse, tighten qualification before spending more.

Mistakes to avoid

• Rewarding sign-ups instead of purchases. You will pay for a list, not a business.
• Asking at the wrong moment. Requesting a referral before the first delivery arrives wastes the strongest moment you will get.
• No new-customer test. Existing customers will re-register under a second number for the discount.
• Cash rewards without verification. Cash attracts organised abuse quickly.
• One-sided rewards for the referrer only. It makes customers feel like commission agents among friends.
• Codes nobody can remember or type. Long random strings kill conversion in a WhatsApp forward.
• Not writing the rules down. Qualification, expiry and abuse terms must be published and linked everywhere.
• Ignoring the referrer after the referral. Confirm the status and the reward, or people stop sharing.

Launch checklist

• Reward structure and values modelled against acquisition cost and margin
• Qualified-referral definition written, including new-customer test and holding period
• Codes and links generated for every existing customer
• Checkout captures the code and a "who referred you?" fallback
• Referral states tracked from share to reward, with an event log
• Store credit posts to the same ledger as loyalty balances
• Fraud checks for address, device, phone and velocity in place
• Refund and cancellation reversal implemented
• Share message written for WhatsApp forwarding and tested on a phone
• Programme rules published, including abuse and appeal terms
• Referrer status updates automated
• Reporting pack defined and a review date set

Conclusion

Referral programmes reward the thing that already grows Nigerian businesses: personal recommendation inside a WhatsApp group. The build is modest — codes, links, a qualification rule, a reward ledger and fraud checks — but the design decisions carry the result. Make the offer two-sided, pay only on a qualified purchase, ask right after a good delivery, give people words they can forward, and measure referred customers against your other channels before you scale the reward.

If you want referral tracking that works across your website, WhatsApp and in-store sales without manual reconciliation, Linestech builds referral, loyalty and customer data systems for Nigerian businesses.

Frequently asked questions

What is the difference between a referral programme and an affiliate programme?

Referral programmes reward ordinary customers for bringing people they know, usually with small, occasional rewards. Affiliate programmes reward publishers, creators or marketers for driving volume, usually with ongoing commissions, contracts and tax considerations. Start with referral; add affiliates only when you can manage them properly.

Should I reward the referrer, the new customer, or both?

Both, in most consumer businesses. A two-sided reward turns the share into a gift rather than a request, which removes the social awkwardness that stops many people from sharing at all.

How do I track referrals that happen by word of mouth?

Add a "how did you hear about us?" field at checkout with an option to enter a referrer's code or phone number, and keep a short manual claim window afterwards. Assume some referrals will never be tracked and treat the tracked number as a floor, not a total.

Is cash or store credit the better reward?

Store credit costs you margin rather than cash, returns as a future purchase and attracts less abuse. Cash motivates more strongly and suits high-value products, but needs verification, caps and a longer holding period.

How long should I wait before paying a reward?

Long enough for refunds, returns and pay-on-delivery remittance to settle, typically 7 to 14 days after delivery. For subscriptions, paying after the second billing cycle filters out customers who were only ever taking the introductory discount.

Can a service business use a referral programme?

Yes, and often more effectively than a shop, because service choices depend heavily on trust. Qualification is usually the completed first job rather than an order, and rewards can be a discount on the next service or a credit against a maintenance plan.

Will a referral programme hurt my margins?

Only if the reward exceeds the value of the customer it brings. Model the combined reward against gross margin on the first order and against your existing acquisition cost, cap exposure per referrer, and review after one purchase cycle using real data.

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.

• Nigeria Data Protection Act 2023 — NDPC
• Baymard Institute e-commerce UX research
• Paystack developer documentation