Build specification · August 2026
The strategy, extracted for development
No competitive teardown, no benchmarking case — just what to build, and the constraints that shape it
Companion to “Beyond the Reward Hub”
Contents
One-page brief
The operator can lend — so the program can reward good behaviour with cheaper, faster credit, not just bigger discounts. Six layers deliver that, ordered by how often each one fires.
Primary metric
Engagement — DAU/MAU stickiness, per the MFB's own stated priority. Retention is downstream of sustained daily engagement.
The one guardrail
Engagement bought with small extrinsic rewards spikes in month one and fades by month three.
The fix, already designed in
Pair the highest-frequency layers with a durable hook — Trust Score — that no competitor can outbid with a marginally better daily payout.
One-page brief
What to track, same cohort
Whether customers with rising Daily Habits and Moments engagement also show rising Trust Score movement.
Both series need to be queryable by cohort from day one — this is an instrumentation requirement, not a reporting nice-to-have.
The signature to act on
Engagement climbing while Trust Score stays flat — reward-chasing rather than habit formation.
Fixable by strengthening the Trust Score payoff, but only if it is caught here rather than later as churn.
Engagement up, Trust Score flat, is the early warning — not something to notice once it has become churn.
01
Eight rules for settling build trade-offs without reopening the strategy.
Design principles
Credit is the currency, not just coupons — progression feeds a visible, weekly-recalculated Trust Score.
Give it a purpose bigger than discounts — literacy, small-business growth, community impact.
Reward the moment, not the month — instant and visible, never a multi-week draw.
Design for redemption, not breakage — a program profitable on unredeemed points depends on disengagement.
Keep the base layer boringly simple — real cashback, credited automatically, no ledger to interpret.
Reward the KYC upgrade itself — front-load a bonus for clearing biometric verification and Level 0.
Fund it with partners wherever possible — elevated rewards run through brands already spending on acquisition.
Make daily engagement free, not compulsory spending — zero-cost actions only.
Level 0 is capped at Rs. 25,000 a day — confirmed, not an estimate — which structurally blocks meaningful engagement before a customer even reaches the program. Principle seven turns that ceiling into the first reward moment.
02
Six layers, ordered by trigger frequency — how often each one gives a customer a reason to open the app.
Program architecture
| Layer | What it is | Cadence | Build note |
|---|---|---|---|
| Daily Habits | Streaks, fraud-awareness and literacy quizzes, light puzzles. The only layer rewarding showing up and learning, not transacting. | Daily, no transaction required | Full mechanics in section 03 |
| Moments | Instant transaction-triggered mini-reward — spin, reveal, small bonus — max once a day, credited instantly. | Up to daily, per qualifying transaction | Spendability gated by Cooldown A |
| Foundation | Automatic cashback on everyday transactions, real-time, with no points ledger. | Every transaction | Default experience; no opt-in complexity |
| Trust Score | Score from transaction volume, bill punctuality and savings behaviour. Unlocks faster approval, higher limits, better pricing. | Weekly recalculation | Needs credit-risk model integration — Phase 2 |
| Marketplace & Purpose | Partner-funded voucher marketplace plus the purpose track — literacy badges, small-business milestones, community impact. | Periodic / aspirational | Partner-funded, not core margin |
| Onramp | One-time bonus for completing biometric verification and clearing Level 0. | Once, at KYC upgrade | Sequence around Cooldown B |
Program architecture
| Tier | Qualifying behaviour | Foundation benefit | Trust Score unlock |
|---|---|---|---|
| Seed | Account opened, first three transactions | Welcome cashback credit | Baseline score established |
| Sprout | Regular monthly transacting — bills, transfers, top-ups | Elevated cashback on two to three transaction types | Score visible; first loan pre-qualification signal |
| Root | Sustained three months or more, on-time bill payments | Partner marketplace access unlocked | Faster approval track; modest limit increase |
| Harvest | High, consistent activity plus a savings product held | Priority support; increased Moments reward rate | Preferential pricing on qualifying loan products |
| Legacy | Top-decile activity 12 months or more, or literacy and small-business milestones | Full marketplace and purpose-track benefits; invitation perks | Best available pricing and limits; fast-track for pilots |
Naming and thresholds are the least important part of this section. The layering — onramp, foundation cashback, credit-linked trust score, instant moments, free daily habits, partner-funded purpose — is what to protect.
03
Streaks first, quizzes second, games last. Every trigger is a free action.
Daily Habits · 4.1
Reward curve
| Day | Baseline | Milestone | Total |
|---|---|---|---|
| Day 1 | 1 pt | — | 1 pt |
| Day 7 | 1 pt | +2 pts | 3 pts |
| Day 14 | 1 pt | +4 pts | 5 pts |
| Day 21 | 1 pt | +6 pts | 7 pts |
| Day 30 | 1 pt | +9 pts | 10 pts + tier credit |
Build requirements
One free streak freeze per calendar month. A single missed day must not reset progress.
No push notification shaming a broken streak. One gentle, neutral reminder is acceptable; guilt-based re-engagement copy is not.
Trigger: app open plus a tap. Zero-cost by design — see principle eight.
Daily Habits · 4.2
Two to three questions per session, rotating content, one to two points per correct answer. The explanation is shown after every answer, right or wrong. Starter bank below — expand it, weighted toward branchless-banking fraud patterns.
“Someone calls claiming to be from your bank and asks you to read out the OTP you just received. Should you share it?”
No — banks never ask for your OTP over a call.
“A message says you've won a prize and must pay a small fee via QR code to claim it. Is this safe?”
No — a legitimate prize never requires payment first.
“Is it ever okay to give your MPIN to an agent to complete a transaction for you?”
No — agents never need your MPIN.
“Before scanning a merchant's QR code, should you check the displayed name matches the merchant you're paying?”
Yes — QR-code swaps are a real fraud pattern.
Daily Habits · 4.2–4.3
The metric to instrument
Quiz completion correlated against downstream fraud-dispute rates, by cohort.
If the correlation holds, this layer is partially self-funding through avoided dispute and fraud-loss costs — report it to Risk and Finance that way, not just as an engagement number.
Light games — defer, don't build first
Sequenced into Phase 4, after streaks and quizzes have proven the habit loop — however tempting it is to build the fun part first.
Tie them to the purpose narrative — a savings-vocabulary word puzzle, a “grow your plot” mechanic — not generic arcade content unrelated to the brand.
04
Confirmed, not theoretical. Each one changes what the UX is allowed to imply.
Regulatory constraints · 5
Cooldown A · funds-received hold
2 hours
Cash-outs, online purchases and top-ups against newly received wallet funds are held two hours — SBP, confirmed. Any instant Moments reward is itself “funds received”.
Build requirement: credit it visibly and instantly, but the UI must say it becomes spendable after the standard window. Do not imply instant usability.
Cooldown B · security-action hold
2 hours
A broader freeze on all debit transactions, triggered by registration, new-device registration, email change, or MPIN change — confirm the target MFB's equivalent rule before build.
Build requirement: if the Onramp flow registers a new device, expect a blanket debit freeze straight after. Celebrate the bonus, but sequence messaging accordingly.
Regulatory constraints · IFRS 15
Liability = Outstanding Points × (1 − Breakage Rate) × Cost Per Point
Why two models
Daily Habits generates many small habitual points that pool up. Moments generates occasional larger instant rewards. Different volume and redemption shapes.
Launch gate
Breakage tracking instrumented from day one; Finance and Audit sign-off on the liability model required before launch.
Regulatory constraints · 5
Level 0 ceiling
Rs. 25,000
per day — confirmed against the source terms
This is why the Onramp bonus exists: a Level 0 customer is structurally capped out of meaningful engagement with the rest of the program until they clear it.
Shariah framing, if an Islamic-window variant ships
Frame rewards as discretionary institutional generosity — hiba — tied to real transactions. Never as a promised or rate-benchmarked return.
Chance-based mechanics are workable where the entry is a transaction already happening for its own purpose, not a discrete paid stake.
Get named Shariah board sign-off on the final mechanic — not the marketing copy — before launch.
05
A funding rule per layer, and four phases over eighteen months.
Funding & roadmap · 6
Onramp
Fixed, small, one-time cost per KYC upgrade — budget it as customer acquisition cost.
Moments
Fixed marketing budget, capped like a CAC allocation — not an open-ended prize pool.
Foundation
The bank's own margin, kept deliberately modest — its job is trust-building, not maximum reward value.
Daily Habits
Cheapest layer: small non-cash points, largely redeemed into the partner-funded marketplace. The fraud quiz is trackable as partially self-funding.
Trust Score
Self-funding: better pricing for lower-risk, higher-engagement customers reduces cost-of-risk enough to offset the concession.
Marketplace & Purpose
Funded predominantly by merchant and brand partners paying for elevated visibility and earn rates.
Funding & roadmap · the deposit angle
Every rupee of Average Daily Balance that a tier pulls into savings and current accounts is low-cost retail funding, reducing blended cost of funds — measured in basis points.
Applies to the Root and Harvest tiers. Track it jointly with Treasury, not Product alone.
Funding & roadmap · 7
| Phase | Timeline | Ship |
|---|---|---|
| Phase 1 | 0–3 months | Onramp bonus, daily check-in streaks, Foundation cashback and the basic Seed–Root ladder, all on existing transaction data. Breakage tracking instrumented from day one. |
| Phase 2 | 3–9 months | Trust Score built and validated against existing credit-risk models, piloted with a controlled cohort. Fraud-awareness and literacy quiz content library shipped. |
| Phase 3 | 6–12 months, parallel to Phase 2 | Moments — instant transaction-triggered rewards — shipped. First three to five funded Marketplace partners onboarded. |
| Phase 4 | 12–18 months | Purpose-linked track launched. Lightweight purpose-themed puzzle layer shipped — only once the habit loop is proven. Marketplace scaled. Full Shariah-compliant variant if applicable. |
06
How we will know it is working — and what we are deliberately not building in v1.
Measurement · 8
Engagement — headline
DAU/MAU stickiness, weekly — the north star, and a leading churn indicator two to three months ahead of revenue.
Streak completion at Day 7 / 14 / 30.
Quiz completion and correct-answer rate.
Session frequency by cohort — one ratio can hide a power-user core propping up the average.
Durability — mercenary or real
Daily Habits and Moments engagement against Trust Score movement, same cohort.
Trust Score to repayment and default correlation, against a control cohort.
Retention delta, participants versus non-participants — segmented by whether engagement was Daily-Habits-led or Trust-Score-led.
Program economics
Redemption rate — target sustained and rising, never minimised as a cost saving.
Partner-funded share of total reward value, target rising over time.
Points liability as a share of program revenue — modelled separately per layer, tracked quarterly with Finance.
Risk · 9
| Risk | Mitigation |
|---|---|
| Quietly becomes a discount-only clone, losing the credit-linked differentiator during build | Protect Trust Score as a funded, tracked workstream from day one — not a Phase 2 item that slips |
| Regulatory or Shariah non-compliance in reward structuring | Compliance and the Shariah board involved in mechanic design, not just marketing review |
| Points liability understated, creating an audit finding | Joint Finance and Audit liability model built and reviewed before launch |
| Cannibalising loan margin through overly generous Trust Score pricing | Pilot pricing and limit changes with a controlled cohort; risk-model validation as a launch gate |
| Execution quality — support, redemption reliability — undermines an otherwise sound design | Treat redemption reliability and support responsiveness as core product KPIs from day one |
| One big-bang release delays time-to-market | Phase 1 ships on existing data, independent of the more complex Trust Score work |
Scope · 10
Light games and puzzle mechanics — deferred to Phase 4, after the habit loop is proven with streaks and quizzes.
A quantified financial model — exact NFI from a premium tier, exact basis-point impact on cost of funds. Needs the MFB's real cost-of-funds and adoption assumptions; do not substitute invented placeholders.
The full Shariah-compliant variant — Phase 4, with board sign-off on the mechanic before the phase starts, not retrofitted after.
Exact account-tier thresholds above Level 0 — pull them from the MFB's live Schedule of Charges before finalising the Onramp bonus.
Named here so scope-creep has something concrete to point back to.
What to protect through delivery
The names and thresholds are workshop material. The layering is the part worth protecting, whatever the tiers end up being called.