Core modernisation in 24 months: a comparison of four playbooks.
Banking core modernisation is rarely blocked by technology alone. It stalls when the chosen migration playbook does not fit the bank's product complexity, regulatory posture, and appetite for dual-run risk. The question is not which playbook is fashionable; it is which one can finish under real constraints.
Most programmes fail because the migration logic is mismatched to the institution.
The strongest predictor of success in the programmes we reviewed was not vendor selection. It was whether the bank had chosen a migration playbook aligned to its product architecture, country variation, and tolerance for parallel operations. Too many programmes adopt a pattern because it worked elsewhere under very different conditions.
Banks that finish tend to decide early where they are willing to carry duplication, how much business variation they will rationalise before migration, and what evidence regulators will need to trust the move. Those answers usually point toward one playbook more clearly than executives expect.
Choosing a playbook means deciding where complexity lives during the transition and which trade-offs the organisation can realistically absorb.
How the four playbooks compare in practice.
Strangler works when product domains can be teased apart cleanly
It offers strong control and visible progress, but it demands disciplined API contracts, event models, and a willingness to run mixed-state architecture for longer.
Big-bang only fits narrow complexity bands
It can reduce long transition periods, but the evidence burden, rehearsal intensity, and customer-risk profile make it unsuitable for most large multi-market banks.
Sidecar is useful when the bank needs rapid product change before full migration
A sidecar approach can modernise selected journeys quickly, though it risks becoming a permanent integration layer if roadmap discipline slips.
Greenfield is strongest when the bank is willing to simplify first
Creating a new target core for selected products or markets can accelerate change, but only if leaders are prepared to retire legacy variation and policy exceptions.
The decisions to make before the programme hardens.
Quantify product and market variation
Map where local product logic, servicing rules, and regulatory differences genuinely drive value versus where they are just inherited complexity.
Choose the tolerated transition burden
Decide how much dual-run cost, integration duplication, and reconciliation effort the organisation can carry for 12 to 24 months.
Align the playbook to regulatory evidence needs
Rehearsal design, rollback proofs, and customer-protection controls should shape the migration approach before platform selection begins.
Protect the modernisation from adjacent change
Large programmes finish more often when major product overhauls, market launches, and policy redesigns are sequenced intentionally around the migration path.
Where each playbook tends to fit best.
Universal banks often favour strangler or sidecar paths
These institutions typically need to preserve service continuity while gradually reducing complex market-specific dependencies.
Digital challengers can move faster with greenfield constructs
Cleaner product sets and less historical variation make it easier to design a new target state without dragging every legacy rule forward.
Regional product lines occasionally justify selective big-bang moves
Smaller scope, limited product families, and strong rehearsal capacity can make a concentrated cutover economically sensible.
The migration pattern is the strategy. Once you choose it, you have already decided where the pain and the proof obligations will sit.Marcus Hale - Banking & Financial Services Partner, Tata Consulting Services
Explore adjacent insights from the same research stream.
Unified commerce: the data plumbing nobody warned you about.
Why unified-commerce programmes fail first in the integration layer, and the data, inventory, and event patterns that stand up under peak retail demand.
Workforce studyWhat 8,400 frontline workers told us about AI augmentation.
Findings from our frontline workforce study on where AI actually helps, where it creates friction, and what workers need before they trust augmentation at scale.
ESG & sustainabilityScope 3 reporting: the data engineering challenge nobody is talking about.
How organisations are building the data, supplier, and evidence pipelines needed for credible Scope 3 reporting when the source data is incomplete, inconsistent, and distributed.