"EA Case Study: Retail Bank Modernization Program"
A hypothetical end-to-end enterprise architecture program for a mid-size bank, from capability assessment and target architecture to governance, roadmap, funding and results.
[!ASSUMPTION] Hypothetical case study combining common patterns from banking modernization programs. Figures are illustrative.
Situation
A mid-size retail bank (6 M customers) faces fintech competition. Its estate: mainframe core banking, 420 applications, three CRMs from past acquisitions, nightly batch integration, two on-premises data centers, and a 9-month average time to launch a product. Regulators raised findings on operational resilience and data quality.
Approach (ADM-informed, agile execution)
1. Vision and capabilities (8 weeks)
- Executive goals: product launch < 3 months, digital sales 50% of new accounts, cost-income ratio −8 points, resolve regulatory findings.
- Capability heat map: red for Customer Onboarding, Product Configuration, Customer Data Management, Payments Processing; amber for Lending.
2. Portfolio assessment (6 weeks)
- TIME analysis: 85 applications to eliminate (duplicates, unused), 60 to migrate, 40 to invest.
- Three CRMs → one SaaS CRM (repurchase).
3. Target architecture
flowchart TB
CH[Digital channels<br/>mobile, web, branch] --> API[API platform<br/>gateway + BFFs]
API --> DOM[Domain services<br/>onboarding, accounts, payments, lending]
DOM --> LEDGER[Cloud-native core ledger<br/>new products]
DOM --> ACL[Anti-corruption layer] --> MF[(Mainframe core<br/>existing products)]
DOM --> EV[[Event streaming platform]]
MF -- CDC --> EV
EV --> MDM[Customer MDM] --> DOM
EV --> LAKE[(Lakehouse: reporting,<br/>risk, regulatory)]
CRM[SaaS CRM] <--> EV
subgraph Platform
LZ[Cloud landing zone] --- OBS[Observability] --- SEC[Security & IAM]
end4. Principles and governance
- 10 principles (buy for commodity, API-first, data ownership, cloud-managed first, secure by design).
- Lightweight ARB with triggers; ADRs mandatory; landing-zone policies as code.
5. Roadmap
| Wave | Timeline | Outcomes |
|---|---|---|
| Foundations | Months 0–9 | Landing zone, API platform, streaming + CDC from mainframe, identity modernization |
| Customer | Months 6–18 | MDM golden record, digital onboarding, CRM consolidation |
| New core | Months 9–24 | New savings products on cloud-native ledger |
| Migration | Months 18–48 | Migrate existing products off mainframe; decommission |
| Data & AI | Months 12–36 | Lakehouse, regulatory reporting automation, fraud and personalization models |
6. Funding and organization
- Product-aligned teams per domain; platform team for landing zone, APIs and streaming.
- Funding per value stream with quarterly reviews against KPIs.
Results after 24 months (illustrative)
- Product launch time reduced from 9 months to 10 weeks for new savings products.
- 85 applications retired; licence savings funded 30% of the program.
- Digital onboarding from 5 days to 12 minutes for 80% of customers.
- Regulatory findings on data quality closed via MDM and lineage.
Lessons
- Early wins (onboarding, CRM consolidation) built credibility to fund the core migration.
- CDC from the mainframe unlocked value before any code migration.
- Governance worked when automated and fast; the first ARB design (monthly, document-heavy) was replaced within 3 months.
- Decommissioning needed dedicated ownership and executive sponsorship.
Interview questions and answers
What would you do in the first 90 days as EA in this bank?
Understand strategy and pain with executives, build a quick capability heat map and application inventory, identify 2–3 high-value initiatives, set up lightweight governance and principles, and deliver an initial target architecture and roadmap for funding.