Deepening relationships or just subsidising swipes?

Financial Services Rewards Assessment — Full Reference Content

Intent — The Product Engine (3.2x)

FNB eBucks members hold 3.2 products versus 1.4 for non-members. The programme exists to deepen the relationship, not to give away rewards.

Yet 2025 showed the fragility. Tier cuts, partner removals, and forced product uptake triggered backlash. When a programme shifts from rewarding behaviour to demanding behaviour, loyalty erodes.

Intent — Low Tier Recommendation

Define one measurable behaviour your programme exists to drive. Not retention in general. A specific action: increase product cross-sell from 1.4 to 2.0, or reactivate 10,000 dormant accounts in 90 days. Attach a reward to that behaviour and measure only that.

Intent — Mid Tier Recommendation

Your programme has a goal but the link between reward and behaviour is loose. Tighten it. Map each reward tier to a specific product-holding or activity threshold. Remove rewards that do not drive the target behaviour.

Intent — High Tier Recommendation

Strong intent. Now stress-test it. FNB’s 2025 backlash showed that even mature programmes break when they shift from rewarding existing behaviour to demanding new behaviour. Audit whether your programme rewards what customers already do or forces what you want them to do.

Coverage — The Penetration Benchmark (92%)

Capitec enrols 92% of its 25 million clients in Live Better. No sign-up friction. No points. Real cashback into an interest-bearing savings account.

If your programme covers less than half your base, you are running a perk for your best customers, not a growth tool for the business.

Coverage — Low Tier Recommendation

Start with your dormant accounts. 40-60% of accounts sit inactive. A targeted reactivation campaign, R30 fuel voucher for three transactions this month, costs less than acquiring a new customer and proves the model before you scale.

Coverage — Mid Tier Recommendation

You reach more than half your base but not the mass market. Check your channel mix. If you are app-only, you are invisible to customers who bank via USSD or WhatsApp. Add one low-friction channel and measure the coverage lift in 90 days.

Coverage — High Tier Recommendation

Strong coverage. Focus on depth. Capitec’s 92% enrolment means nothing if engagement is shallow. Track monthly active participation, not just enrolment. Set a target: 60%+ of enrolled customers earning or redeeming at least once per month.

Relevance — The Saturation Problem (10.3)

South African consumers juggle 10.3 active loyalty programmes. 82% participate. 77% say programmes influence where they buy groceries.

Points programmes compete with 9 other programmes for attention. Lifestyle rewards that solve daily problems, taxi fare, school fees, airtime, compete with nothing.

Relevance — Low Tier Recommendation

Your rewards do not solve daily problems. Start with one category your customers spend money on every week: groceries, fuel, airtime, or transport. A R10 grocery voucher for maintaining account activity creates more loyalty than 500 points redeemable in 60 days.

Relevance — Mid Tier Recommendation

Your rewards have some daily relevance but the range is narrow. Add one lifestyle category you do not currently cover. If you offer cashback, add fuel. If you offer fuel, add groceries. Each category added reduces the chance of a competitor being more relevant.

Relevance — High Tier Recommendation

Strong relevance. Now personalise. Capitec uses WhatsApp to deliver hyper-personalised offers at 20-40% conversion. Segment your rewards by customer behaviour, not just tier. A customer who buys fuel weekly should see a fuel offer, not a travel deal.

Delivery — The Speed Gap (70%)

Capitec’s WhatsApp campaigns achieve 70% read rates, 40% click-through, and 20-40% conversion. The value is immediate and the channel is where the customer already is.

A monthly statement credit arriving 45 days after the qualifying action is not a reward. It is a reconciliation line.

Delivery — Low Tier Recommendation

Your reward gap is too long. A monthly statement credit does not feel like a reward. Move one reward to instant delivery via WhatsApp or USSD. Even if the value is small, R5 airtime delivered in 60 seconds after a qualifying action, the speed creates the behaviour loop.

Delivery — Mid Tier Recommendation

You deliver within a reasonable timeframe but not instantly. Test instant delivery on one campaign. Compare the redemption rate and repeat behaviour against your standard delivery. The data will justify the infrastructure investment.

Delivery — High Tier Recommendation

Strong delivery. Optimise the channel mix. If you deliver via app but your mass-market customers use WhatsApp, you are fast but invisible. Match the delivery channel to the customer segment, not the internal system.

Measurement — The Dormancy Cost (40-60%)

40-60% of retail bank accounts are dormant or low-activity. These accounts cost money to maintain and generate no fee income.

A reactivation reward tied to a specific behaviour, three transactions this month, a new debit order, is cheaper than acquiring a new customer and proves the programme works.

Measurement — Low Tier Recommendation

You cannot prove your programme works. Start with a simple hold-back test: run one campaign with a reward group and a control group. Measure the difference in the target behaviour after 90 days. One test gives you more proof than a year of redemption reports.

Measurement — Mid Tier Recommendation

You measure activity but not attribution. Add control groups to your next campaign. Isolate the programme’s impact from general marketing. If you cannot prove the reward caused the behaviour, you cannot justify the budget.

Measurement — High Tier Recommendation

Strong measurement. Build a cost model. FNB can show that eBucks members generate more revenue through product holding. Capitec can show campaign conversion at 20-40%. Calculate the exact ROI per reward rand spent and use it to negotiate budget increases.

Cost of Doing Nothing — Intent (3.2 vs 1.4)

FNB eBucks members hold 3.2 products versus 1.4 for non-members. Without clear intent, your programme subsidises existing behaviour instead of creating new value. Every month without a defined behaviour target is a month of reward spend with no measurable return.

Cost of Doing Nothing — Coverage (40-60%)

40-60% of your accounts may be dormant. Each dormant account costs money to maintain and generates zero fee income. A competitor offering a R30 reactivation reward is cheaper than your next acquisition campaign.

Cost of Doing Nothing — Relevance (10.3)

Your customers juggle 10.3 loyalty programmes. If your rewards are not solving a daily problem, you are competing with 9 other programmes for attention you will not win. Every month of generic rewards is a month your competitor’s grocery voucher feels more valuable than your points.

Cost of Doing Nothing — Delivery (45 days)

A reward delivered 45 days after the qualifying action is a reconciliation line, not an incentive. Capitec delivers value instantly via WhatsApp at 70% read rates. Every month you delay delivery is a month your competitor’s instant reward builds the habit instead of yours.

Cost of Doing Nothing — Measurement (R0)

Without attribution, your programme is a cost centre that cannot defend its budget. The first executive who asks for proof of ROI will cut the programme. Every month without a control group is a month closer to that conversation.