Switching FinTech Stack

Description

If the switch is being driven by insolvency or under ninety days of runway, this is the wrong Playbook. Navigating a Cash Crisis assumes urgency this one doesn’t.

Stack switches rarely fail because you picked the wrong vendor. They fail on sequencing.

A Direct Debit mandate in the UK, or an ACH authorisation in the US, that doesn’t get re-established in time. A reconciliation that breaks mid-month because two systems were both live and nobody was cross-checking. A customer’s saved card details that don’t migrate cleanly, so the next renewal silently fails. An old account closed a week too early, before the last mandate had finished redirecting. None of those are vendor problems. All of them are order-of-operations problems, and every one is preventable.

What you’ll have at the end of eight weeks

A new provider fully live at the layer you switched. Every recurring payment, inbound and outbound, continuing without a missed or duplicated instalment across the cutover. A first post-cutover reconciliation that ties out cleanly. The old provider formally off-boarded, data exported, account closed on your terms rather than left open quietly leaking fees. And a rollback plan that existed throughout, whether or not you ever needed it.

Done properly, the whole thing should look almost boring from the outside: nothing broke, nothing was late, and new bank details on next month’s invoice are the only visible sign anything changed.

The six chapters

1. Why Stacks Get Switched, and the Vendor Menu Read Straight — forced versus chosen triggers, and the UK and US vendor menu across bank, payments, accounting and invoicing
2. Mapping Your Dependency Chain — a single-page map of what depends on the layer you’re switching, with payroll flagged as the highest-consequence dependency
3. The Payment Mandate Inventory — every inbound and outbound mandate inventoried before cutover, including the CASS-versus-manual-ACH divergence that changes the whole chapter
4. Building the Parallel-Run and Data Migration Plan — how long to run old and new side by side, what to migrate versus archive, and guarding against bank-detail-change fraud
5. Cutover Week: Running the Switch Live — choosing the date, the cutover sequence, the first 48 hours, and the first post-cutover reconciliation close
6. Closing Out the Old Stack, and the Rollback Plan You Hopefully Didn’t Need — off-boarding, records retention, and retiring the rollback plan

The seven tools

Stack Dependency Map · Payment Mandate Inventory Tracker · Vendor Evaluation Matrix · Parallel-Run Reconciliation Tracker · Cutover Day Runbook · Rollback Plan Template · Off-Boarding and Closure Checklist. Spreadsheets and checklists you build and use as the chapters call for them — nothing beyond a standard spreadsheet package.

UK and US, in one place

Where the mechanics diverge, the divergence is flagged inline at the point it matters. CASS in the UK against manual ACH re-authorisation in the US isn’t a footnote — it changes how an entire chapter runs. Two worked examples run throughout: Thornbury & Co in the UK, Meridian Fulfillment in the US.

Who it’s for, and where to start

The founder, finance lead, or operations lead of a UK or US SMB, roughly £500k–£15m or $750k–$20m turnover, who owns or heavily influences the switching decision. Assumes two to four hours a week across eight weeks.

Four legitimate ways in. Haven’t chosen a new provider yet: work it in order. Already signed and planning the migration: start at Chapter 2 and run through to Chapter 6 — that’s the core sequencing work. Being forced out on a shorter timeline by a vendor exit notice, an account closure, or a processor terminating your agreement: go to the accelerated-timeline note at the end of Part 1, then straight to Chapter 3, which is the highest-priority work regardless of how little time you have. Already mid-migration and something has gone wrong — a missed payment, a reconciliation that won’t tie out, a mandate that didn’t redirect: go directly to Chapter 5 and the Rollback Plan Template. Stabilise first, read the rest after.

What it doesn’t cover

Building a stack from nothing for a brand-new business — Chapter 1’s vendor menu helps, but the rest assumes a live stack with existing dependencies and customers that need to keep working. Multi-entity, multi-subsidiary, or cross-border treasury consolidation is also out of scope; that carries structural complexity the general chapters don’t resolve.

Format

One-time purchase — yours to keep, and worth holding onto for the next switch as much as this one.