A remediation programme can report rapid progress while leaving the underlying weakness intact. Files may be marked complete because a document was obtained, even though ownership remains unclear or updated information never reached monitoring systems. The useful objective is to reduce unresolved exposure and prevent recurrence, with completed-file counts serving as supporting information.

The FCA’s April 2026 CDD review addresses CDD processes and assurance. The Financial Crime Guide provides broader control context. The programme design below is a management framework; applicable law, supervisory commitments and case-specific facts determine required deadlines and interim actions.

Define the population and the defect

Create a reliable inventory of affected relationships. Explain how it was identified and reconcile it to the relevant systems. Record duplicate cases, closed accounts, linked entities and files that moved into another review process. Otherwise the apparent reduction in a queue may reflect administrative movement rather than resolution.

Classify the defect precisely. Missing evidence of identity, unresolved ownership, an outdated activity profile and an unprocessed risk-rating change are different problems. They require different skills, information and completion tests. A single “KYC overdue” category hides those distinctions and makes staffing estimates unreliable.

Record when the weakness arose, when it was discovered and which period may be affected. Remediating the current file may not answer whether historical activity needs assessment. Assign that question explicitly rather than letting it disappear between the project and business-as-usual teams.

Prioritise exposure as well as age

Use a documented triage method that considers the significance of the missing information, customer activity, relevant risk factors and existing controls. Age is important but should not be the only criterion. A newly discovered material ownership gap may warrant attention before an older, lower-impact administrative defect.

Treat uncertain information carefully. Missing data should not automatically place a case into a low-priority category simply because the model cannot score it. Define an escalation route for cases where the lack of information itself affects the ability to judge risk.

Interim measures need separate decisions. Where a restriction, additional monitoring or other action is considered, document its legal basis, rationale, owner and review point. A remediation project should not invent blanket measures without assessing their consequences and the applicable framework.

Make completion criteria observable

For each defect type, state what must be true for closure. This could include verified evidence, a reasoned ownership conclusion, a revised risk assessment, the required approval and confirmation that downstream systems were updated. Distinguish cases that are fully resolved from those transferred for a further decision.

Require a clear exception record where complete remediation is not possible. The record should explain the unresolved issue, the authorised decision and the next review point. Relabelling uncertainty as completion creates a misleading programme result.

Give reviewers examples of satisfactory and unsatisfactory evidence. Calibration sessions can compare decisions on the same case and identify inconsistent interpretations early. Staff should be rewarded for correct decisions and useful escalation, not solely for closing the largest number of files.

A hypothetical backlog reduction

Assume a bank reduces its queue by half in a month. Independent sampling finds that many files contain newly uploaded documents but no recorded assessment of their relevance. Some corrected names have not reached the screening platform.

The programme should determine which completed cases may share these defects, correct the closure standard and reassess the affected population. It should not assume the sampled error rate applies uniformly to every team or period. Stratified analysis can identify where the weakness is concentrated and where wider review is justified.

This example illustrates why quality review needs to run while remediation is underway. Discovering a systematic defect only after the programme closes can require expensive rework and leave management with an inaccurate view of exposure.

Test whether the solution lasts

Monitor new cases entering the backlog and compare their causes with those being remediated. If the same onboarding defect continues to create incomplete files, additional project capacity merely treats the symptoms. Assign ownership of the process fix and verify that it changes new-case outcomes.

Management reporting should distinguish resolved exposure, quality failures, rework, newly discovered cases and remaining uncertainty. Forecast completion using realistic productivity and rejection rates, with sensitivity to complex cases. The programme is ready to close when its evidence supports the closure decision and the normal operating process can maintain the improvement.