Integration reporting green while critical retention falls

92% green means the plan is measuring the wrong thing. It is tracking deliverables completed - documents, systems configured, workshops held - not whether the merged organisation can operate. Turnover concentrated in branch managers and mortgage underwriters is the opposite indicator: it is the loss of exactly the people who hold the process knowledge and customer relationships that 'orderly integration' depends on.

Three specific readings:

• The RAG is activity-based. A plan can be green while the capability to run the business after cutover is falling.
• Voluntary turnover rising from 9% to 22% in a single quarter is not noise. In an acquisition where the acquired brand keeps its identity, it usually signals unresolved status anxiety and unclear futures, amplified by the dual authority that the exception routing reveals.
• 'Not their problem' on the Coastline side is disengagement in the acquiring business - often more dangerous than overt resistance, because it produces no signal until integration hits a dependency.

I would re-baseline the report with people measures alongside delivery measures: retention in named critical roles, clarity of decision rights, adoption of new processes, complaint volumes and handling times, and time-to-competence. Then take the honest score to the integration board. It is probably amber, and telling the regulator it is green is the larger risk.

This is a decision-rights failure presenting as a behaviour problem, and communications will not fix it.

Two branch managers refuse the referral process; three use it but route exceptions to an old area manager who no longer has authority. That tells me the new process has no usable exception route, and the old route has never been formally withdrawn. People are not being disloyal; they are using the only escalation channel that still answers.

What I would do, in order:

• Publish the decision rights: who decides referrals, who decides exceptions, and to what service standard. If the new process has no exception path, add one - a process without exceptions is abandoned at the first difficult case.
• Close the old channel explicitly. The former area manager should redirect requests publicly and in writing, visible to the branch managers. A residual authority that is not withdrawn competes with the new one indefinitely.
• Meet the two refusers individually. Diagnose capability (they do not know how), incentive (the old process suited their branch), or identity (the merger threatens their standing). Each needs a different response; a blanket instruction addresses none.
• Use the three partial adopters as the test. If the exception route works for them, word travels faster than any briefing.

The regulator will find written decisions more convincing than a communication plan.

I would not build a retrospective evidence pack; I would use the deadline to force the decisions that produce evidence.

Define the standard first. 'Orderly integration' means customers receive consistent outcomes, decisions have a named owner, critical knowledge is retained, and problems are found and fixed rather than hidden. Then evidence each one:

• Decision rights: published delegation and logs showing exceptions resolved within a stated time.
• Continuity of service: complaints volumes and handling times per brand, migration error and remediation rates, continuity test results.
• Capability: retention in named critical roles, vacancy coverage, and competence sign-off for staff using the new processes.
• Governance: board minutes that show the 22% turnover and the two refusals being raised, with actions and owners - not a clean dashboard.

The 92% green report is a liability in this context. I would replace it with one that shows issues and remediation, because a regulator is more likely to accept 'here is what is wrong and what we are doing' than a green plan contradicted by retention data.

Ten weeks is enough to fix decision rights and the exception route and to establish honest reporting. It is not enough to reverse turnover, and I would not promise that.