Pilot Proposal · May 2026
Commune-Level Agent Banking Pilot
A proposal for extending financial access to agricultural communities in Binh Dinh
Prepared by: Tran Gian Thu HoMay 2026
Slide 2 — Executive Summary
The Case in Four Lines
Problem
62M rural Vietnamese lack practical access to financial services
Solution
Agent banking network at commune level using existing retail infrastructure
Ask
18-month pilot in 5 agricultural communes in Binh Dinh
Target
40% of target population with active accounts by month 18
≈ $100k USD
Cost per target beneficiary: ~100,000 VND ($4 USD) — competitive with comparable programs in Bangladesh and India.
Slide 3 — The Problem
Distance Is a Policy Failure, Not a Geography Problem
8km
Average distance to nearest bank branch in rural Binh Dinh
45 min
Average wait time per bank transaction
73%
Rural residents using banks only for mandatory transactions
0%
Formal credit access rate among surveyed small farmers
The consequence is not just inconvenience. Farmers holding cash miss savings interest. They cannot access credit at reasonable rates. Emergency expenses become crises. The financial system exists — it is simply inaccessible. This is a distribution problem, not a product problem.
Slide 4 — The Solution
Agent Banking: Proven Model, Unimplemented in Vietnam
Authorize existing retail shops — pharmacies, phone shops, general stores — to provide basic banking services using simple technology and biometric verification. The infrastructure is already there. We are adding a license and a float.
Services
Cash deposit/withdrawal, bill payment, account opening, balance inquiry
Technology
Simple tablet + mobile app + biometric verification
Regulatory model
Bank-led agent network, compliant with SBV Circular 39
Precedent
Bangladesh (bKash), Kenya (M-Pesa), India (Business Correspondents)
Slide 5 — Target Geography
Why Binh Dinh — And How to Select Five Communes
Binh Dinh is the agricultural community I know firsthand. I understand the local livelihoods, retail infrastructure, and relationship between commune officials and residents. Selection criteria for pilot communes: (1) distance from nearest bank branch >5km, (2) population >3,000, (3) existing retail infrastructure with fixed-location shops. Final sites would be confirmed through baseline validation; the profiles below are planning assumptions.
| Candidate Profile | Distance to Branch | Est. Population | Retail Shops |
|---|
| Coastal rice-growing commune | 9.2km | 4,800 | 12 |
| Lowland mixed-crop commune | 7.5km | 5,200 | 8 |
| Foothill farming commune | 11.0km | 3,600 | 6 |
| Market-linked farming commune | 6.8km | 5,400 | 15 |
| Remote smallholder commune | 8.4km | 5,900 | 10 |
| Total | avg. 8.6km | ~25,000 | 51 |
Slide 6 — Agent Selection Criteria
Who Qualifies as an Agent
The agent is the product. Selection criteria are non-negotiable. A weak agent network destroys trust faster than any marketing rebuilds it.
1
Fixed location Existing business with a permanent physical address — not mobile or seasonal.
2
Established operation Operating for 2+ years. Stability signal, not a formality.
3
Digital capable Owner is literate and can operate a smartphone app. Basic training provided.
4
Community trusted Verified by commune officials and local networks. Reputation precedes the role.
5
Cash float commitment Willing and able to maintain minimum 5M VND (~$200) float for daily transactions.
Slide 7 — Implementation Timeline
18 Months from Recruitment to Scale Assessment
Month 1–3
Recruitment & Training
Agent recruitment across 5 communes. Target: 15 agents (3 per commune). 2-day training program. Technology setup and system integration.
Target: 15 agents recruited
Month 4–6
Supervised Soft Launch
Full-time field officer support for each agent. Transaction limits in place. Feedback collection and process refinement. Community awareness events in each commune.
Target: 500 active accounts
Month 7–12
Full Operation
Remove training wheels. Agents operating independently. Marketing push via commune bulletin boards, word-of-mouth incentives, and first-transaction bonuses.
Target: 5,000 active accounts
Month 13–18
Scale Assessment
Evaluate agent profitability, user retention, and satisfaction. Prepare an expansion proposal for other agricultural communities in Binh Dinh and potential regional replication.
Target: 10,000 active accounts
Slide 8 — Budget
2.5 Billion VND Over 18 Months
| Line Item | Amount (VND) |
|---|
| Agent setup — tablets, signage, training | 150,000,000 |
| Agent incentives — 18 months | 900,000,000 |
| Technology platform — licensing | 500,000,000 |
| Marketing & community outreach | 300,000,000 |
| Staff — 1 pilot manager, 2 field officers | 500,000,000 |
| Contingency (10%) | 150,000,000 |
| Total | 2,500,000,000 |
Equivalent to approximately $100,000 USD at current exchange rates. Cost per agent: ~167M VND. Cost per target beneficiary (25,000 people): ~100,000 VND.
Slide 9 — Success Metrics
What Success Looks Like — and How We Measure It
Primary
40% active account penetration
10,000 of 25,000 target population with at least 1 transaction/month by month 18.
Secondary
Distance to access point: 8km → <1km
All 5 communes have at least one operational agent within 1km of 80% of households.
Tertiary
Agent profitability by month 12
Agents cover operating costs from transaction fees without subsidy by the end of the first year.
Monitoring cadence: monthly dashboard (transaction volume, active accounts, agent uptime), quarterly reviews with field officers, biannual user surveys in each commune.
Slide 10 — Risk Mitigation
Known Risks and Planned Responses
| Risk | Mitigation |
|---|
| Agent fraud | Biometric verification on every transaction, individual transaction limits, real-time monitoring dashboard with anomaly alerts. |
| Low adoption | Community outreach via commune leaders and women's unions. First-transaction bonus (50,000 VND). Targeted messaging in local dialect. |
| Agent dropout | 3-month probation period before full activation. Performance bonuses at 6- and 12-month marks. Pre-screened waitlist of backup agents per commune. |
| Regulatory change | Formal partnership with licensed bank from day 1. Legal review of all materials against SBV Circular 39. Quarterly compliance check with SBV liaison. |
“This pilot is shovel-ready. The question is not whether agent banking works — it is proven in Bangladesh, Kenya, and India. The question is whether Vietnam will deploy it before the rural-urban divide becomes permanent.”
— Tran Gian Thu Ho, May 2026
