What is the group actually supposed to look like?
Every acquisition lands somewhere. If that somewhere is whichever CRM the last deal happened to use, the group never becomes one business — it stays a shelf of firms with a shared logo. The architecture decision gets made by default, one deal at a time, by whoever is closest to the integration.
One target, and a reason to move onto it.
A target operating model with the client and adviser experience sitting on top of the core rather than inside it — so when the core changes, only the integration layer is reworked. Advisers keep the same screens; clients see nothing. One data model and one golden source that every acquisition maps to once.
Designed AI-native, meaning the operation can be measured and automated later without being rebuilt — but the decision in front of you is an architecture decision, not an AI one.
Four things get defined.
Data model
The golden-source schema your data lands on — completeness, revenue integrity and provenance built in from day one, not bolted on after go-live.
Governance rails
The evidence trail and control points any automation will need to run against, whoever builds it — so there is something to check against later.
Migration path
The sequenced route from current state to target state that Build will execute — scoped, scored and priced before a single record moves.
Ownership model
Who holds the architecture after we leave — documented so it's your asset, not a dependency on us or any other vendor.
Outputs, not activities.
A spine, not a shelf
One data model, one target, one landing playbook — a spine that makes deal fifteen cheaper than deal five.
A sequence
What to change first, and what can wait — scoped against the estate as it actually is, not as documented.
No repeat integration design
No fresh integration design per acquisition — the landing playbook your next three deals can follow.
One fixed-price first step.
Target operating model assessment
Fixed price, price on request. You get the target architecture, an honest gap against what you run today, and a landing playbook your next three deals can follow.
Plain answers, no dodging.
We already have an architecture function.
Then this is the map they've been asking for, built from the estate as it actually is rather than as documented.
We're mid-platform-selection.
Good. This is the work that tells you what to select against.