The problem: programmes that were on plan until they were not
A large remediation programme rarely announces its difficulty. The early months tend to look reassuring. Population sizing is signed off. Vendors are appointed. Analysts are hired. The first tranche of files is processed. The MI pack shows steady throughput. Six months in, the throughput plateaus, the QA pass rate begins to slip, and the programme discovers it has been measuring the wrong thing. Nine months in, the second line raises a concern that the definition of a completed file has been drifting; the operations lead disagrees; the Board pack begins to lose coherence. Twelve months in, the programme is behind plan, above budget and under supervisory attention.
That trajectory is not inevitable. It is the predictable consequence of a small number of specific decisions that were, or were not, taken in the first ninety days.
The evidence: what recent enforcement remediation has taught the market
The five failure modes we see repeatedly
Good practice and poor practice in remediation design
Business impact: the cost of drift
A remediation programme that drifts by six months on throughput typically extends by twelve months in total once the rework of the drifted period is factored in. Where the supervisor is engaged, the cost of the additional oversight (skilled-person work, additional testing, remediation of the remediation) frequently exceeds the incremental programme cost. In every case in my direct experience, the aggregate cost of drift has been between two and five times the cost of the specific investment that would have prevented it.
Practical solutions: what the first ninety days should look like
The Claritas approach
We are usually asked to help either at the beginning of a programme, to design it in a way that will not drift, or at month nine, when it already has. The interventions differ. At the beginning we work with the client on the five items above and, once the programme is running, step back to a monthly independent oversight role. In month nine, the work is diagnostic first: we take a sample of completed files, we take a sample of QA-passed rejections, we read the last three Board packs against the last three programme operating committee minutes, and we produce a note that names the specific decisions that would need to be reset for the programme to recover.
In either case the deliverable is short. Programmes rarely fail for lack of paper. They fail for lack of a small number of decisions being taken in the right order. Our job is to name those decisions and to give the Board the confidence to take them.
- 01A written definition of a completed file exists, is endorsed by the second line, and is the same document a QA reviewer, an analyst and an external adviser would apply.
- 02The QA divergence rate between first line and second line is reported as a headline KRI and is on a trend the Board can defend.
- 03The Board pack opens with residual risk position, not with productivity.
- 04One accountable executive, one weekly operating committee, one monthly Board format. Every material decision is taken in one of the three.
- 05An independent readiness view is commissioned at month three, month nine and at the point of programme closure.
- Financial Conduct Authority, Final Notice: Starling Bank Limited, October 2024 (referenced remediation obligations).
- Central Bank of Ireland, Enforcement action: Coinbase Europe Limited, December 2024.
- SEPBLAC, Annual Report 2025, remediation and periodic review findings.
- Wolfsberg Group, Statement on effective monitoring for suspicious activity (relevant to remediation QA design).
