Operating model design after the Kestrel–Ardley merger

The actual problem is not organisational shape; it is duplicate value streams. Four account managers and onboarding at 14 versus 31 days describe a firm with two inconsistent ways of delivering the same client outcome. Any structure that leaves those flows duplicated will fail, whatever the reporting lines.

Functional model - strengths: clear professional homes, single standards per discipline, fastest visible headcount synergy, simpler cost accountability. Weaknesses here: it optimises each function separately, and the customer experience - the thing clients are already complaining about - has no owner. With two clashing cultures, functional consolidation also concentrates the political fight into a few very visible appointment decisions.

Value-stream / segment-aligned model - strengths: one end-to-end owner per client outcome, natural basis for harmonising onboarding and relationship management, clearer line of sight from work to client value. Weaknesses: duplicates specialist capability across streams, slower to deliver synergy, and requires mature product/process ownership that this firm does not yet have (11 product committees suggest it does not).

Recommendation: a hybrid, sequenced. Organise client-facing value streams by client segment (not geography) so the four-account-manager problem has one owner, and consolidate enabling functions centrally. But sequence it: first stabilise the capabilities the model depends on - a single client data/onboarding capability, a harmonised risk and product-approval process, and a defined process-ownership model to replace the 11 committees. Restructuring before those exist just moves the duplication into a new chart.

Trigger to revisit: if turnover in operations breaches roughly 30% or key institutional clients formally signal intent to leave, the pace of structural change is the wrong lever and integration should slow to capability work.

Why it is a value-stream risk: operations staff execute the onboarding and settlement flows. With two legacy processes, they are being asked to run two sets of procedures, two sets of exception rules and two reporting lines' worth of priorities. 24% turnover means the tacit knowledge that makes a 31-day onboarding work at all is walking out of the door. That is a capability loss in the exact flow the merger is supposed to harmonise, and it shows up later as error rates and client escalations - not as a 'morale issue'.

What to do, in order:

1. Diagnose before designing. Identify which roles sit at the seam between the two legacy processes and which knowledge is undocumented. Protection of critical roles is a design constraint on the restructure, not an afterthought.
2. Make the target explicit and finite. 'One onboarding process by Q3 for institutional clients, 10 working days' is something staff can act on. 'Be one firm' is not.
3. Use visible, small wins over announcements. Harmonising one product line's onboarding gives the two cultures a joint achievement and a concrete object to argue about, which is more productive than abstract cultural workshops.
4. Address the loss narrative. Merger survivors need to know what happens to their specialism; the fear is usually about status and expertise, not about the logo.
5. Instrument it. Track turnover by seam role, overtime, error/rework rates and escalation volume - the same measures the value stream uses, so the people signal is read as an operational signal.

Judgement note: the integration director should resist a big cultural programme running ahead of structural clarity. In a merger at month eight, ambiguity is the driver of attrition; clarity of roles and processes reduces it faster than engagement events.

Treat the target as a hypothesis with a testable structure rather than a number to be defended.

1. Decompose the £90m. How much is headcount, how much is vendor/spend consolidation, how much is revenue synergy (cross-sell, retention)? Revenue synergy in an asset manager depends on client consent and investment performance - it is the least controllable element and the most commonly over-counted. A target where 40% is revenue synergy is a different risk profile from one where 80% is cost.

2. Check delivery rate against shape. £22m in eight months sounds strong, but synergy curves are typically back-loaded: the easy facilities and procurement savings come first, and the hard ones - single risk platform, single onboarding, one product committee - come later and cost more. Ask which £22m has been banked and whether it is run-rate or one-off.

3. Test dependency on the operating model. If the remaining £68m assumes a fully integrated process and platform estate, then the synergy target is contingent on the target operating model, which is not yet decided. That dependency should be stated in the business case rather than buried.

4. Look for dis-synergy. Attrition at 24% in operations and threatened client departures are costs against the target. A credible case nets them off.

5. Check the measurement basis. Is 'delivered' measured against a pre-merger baseline that both firms agree on? If the baseline is contested, the target cannot be verified.

Conclusion framing: report a range with confidence levels and the specific decisions that unlock each tranche, rather than a single figure. That gives the board something to govern.