KSh 4.5 Billion in Points That Are Not Driving Behaviour
Safaricom is the dominant mobile operator in Kenya. It serves over 40 million subscribers and operates M-Pesa, the most successful mobile money platform in Africa, processing KSh 20.2 trillion annually for approximately 60 million users.
Bonga Points is Safaricom's loyalty programme. Customers earn points on airtime purchases, M-Pesa transactions, and data bundle purchases. Points can be redeemed for airtime, devices, merchandise, and partner offers.
The problem is that KSh 4.5 billion in Bonga Points sit unredeemed on Safaricom's balance sheet. That is not a sign of a healthy loyalty programme. That is a growing financial liability that is not generating behaviour change.
The Expiry Ruling
Safaricom attempted to introduce point expiry to manage the liability. Customers who did not redeem within a specified period would lose their points.
A Kenyan court blocked the move. The ruling: Bonga Points, once earned, belong to the customer. They cannot be unilaterally expired.
The immediate effect was a 34 percent surge in redemptions. Customers who had ignored their points for months or years suddenly redeemed. Not because the programme became more attractive. Because they feared losing what they had.
That 34 percent surge is the most important data point in the entire Bonga Points programme. It proves that customers value the points. But they do not value them enough to redeem under normal conditions. Only the threat of loss triggered action.
What This Reveals About Points-Based Loyalty
Points are a deferred promise. Earn now, redeem later. The gap between earning and redeeming is where loyalty programmes lose their power.
When the gap is too long, customers forget. When the redemption options are not relevant, customers ignore them. When the points feel too small to matter, customers accumulate and do nothing. And every unredeemed point sits on the balance sheet as a liability.
Safaricom's KSh 4.5 billion is the clearest example of this dynamic at scale. The points were issued. The behaviour that earned them happened in the past. The future behaviour they were supposed to drive never materialised.
Compare this to instant, guaranteed rewards:
| Dimension | Points (Bonga) | Guaranteed Instant Rewards |
|---|---|---|
| Value timing | Deferred (earn now, redeem later) | Immediate (value on every transaction) |
| Behaviour trigger | Requires active redemption decision | Automatic, no customer action needed |
| Balance sheet impact | Growing liability (KSh 4.5B) | Zero liability (reward delivered at point of action) |
| Customer perception | "I have points somewhere" | "I got value just now" |
| Engagement driver | Loss aversion (expiry threat) | Positive reinforcement (guaranteed value) |
| Legal risk | Court can block expiry | No expiry issue (reward already delivered) |
| Data value | Redemption data only (sparse, delayed) | Transaction-level behavioural data (real-time) |
The fundamental difference: points create a future obligation. Instant rewards create a present-moment behaviour loop. One accumulates liability. The other accumulates data.
The M-Pesa Opportunity
Safaricom processes KSh 20.2 trillion through M-Pesa annually. That is approximately US$155 billion. Sixty million users. The transaction data covers airtime top-ups, bill payments, person-to-person transfers, merchant payments, savings, and loans.
This is the largest behavioural data set in East Africa. And it does not have a structured rewards layer.
M-Pesa users currently earn Bonga Points on some transactions. But the reward is deferred, generic, and disconnected from the behaviour that generated it. A customer who pays their electricity bill via M-Pesa and a customer who sends money to a family member earn points in the same way. There is no differentiation, no personalisation, no behavioural nudge.
What a rewards layer on M-Pesa could look like:
Transaction-linked instant rewards. Pay your electricity bill via M-Pesa and receive a KSh 50 airtime bonus instantly. Buy a data bundle above KSh 500 and receive a grocery voucher. Send money to three different people in a week and receive a transport credit. The reward is linked to the specific behaviour, delivered instantly, and funded by the value of the data generated.
Merchant-funded rewards. M-Pesa Lipa Na M-Pesa merchants co-fund rewards for their customers. A restaurant offers a KSh 100 discount on the next visit for customers who spend above KSh 1,000. A fuel station offers a KSh 50 bonus for every third fill-up paid via M-Pesa. The merchant pays for the reward. Safaricom provides the delivery rail. The customer gets instant value.
Tenure and frequency rewards. A customer who has been active on M-Pesa for 12 consecutive months receives a higher reward tier than a new user. This creates switching cost that a competitor cannot replicate with a sign-up bonus.
Partner-funded lifestyle rewards. Grocery vouchers, fuel credits, school fee contributions, entertainment vouchers. Funded by brand partners who want measurable access to identified M-Pesa users. Delivered via the M-Pesa app, SMS, or USSD.
The MTN South Africa Parallel
MTN South Africa's prepaid subscriber base declined for the first time in FY2025, down 0.7 percent. Prepaid revenue fell 2.3 percent, accelerating to negative 3.9 percent in Q4. Voice revenue dropped 8 percent in Q4.
The pattern is the same across African telcos. Prepaid is under structural pressure. Price competition is intense. Customers hold multiple SIMs and shift usage based on deals. Airtime and data rewards alone are not enough because every competitor can match them.
Lifestyle rewards, groceries, fuel, transport, dining, create value that a competing network cannot replicate. A guaranteed grocery voucher on every fifth recharge is unique to the network that offers it. A competitor's data deal is not.
Safaricom has a structural advantage that no other African telco has: M-Pesa. The payment rail, the merchant network, the transaction data, the customer identity. The missing layer is a rewards engine that turns all of that infrastructure into retention.
The Liability Question
KSh 4.5 billion is the current number. If the programme continues to issue points that customers do not redeem, and the court has ruled those points cannot expire, the liability grows every quarter.
There are three paths:
Path 1: Make redemption more attractive. Improve the rewards catalogue. Add lifestyle categories. Reduce the minimum redemption threshold. Make redemption instant and automatic. This reduces the liability by converting dormant points into active rewards.
Path 2: Transition to guaranteed instant rewards. Stop issuing points. Start delivering instant value on every qualifying transaction. The liability stops growing. New behaviour is rewarded in real time. The existing point balance is honoured through a redemption window with enhanced value.
Path 3: Do nothing. The liability continues to grow. The court ruling prevents expiry. The balance sheet absorbs the cost. Customers continue to accumulate points they do not use. The programme remains a cost centre, not a growth engine.
The choice is not complicated. It is a question of when, not whether.