The Procedure Nobody Followed
A wealth management firm issued a revised onboarding procedure eight months ago. It requires two independent checks of customer identity documents, a documented source-of-funds rationale for any account above a set value, and manager sign-off before activation. An internal review of 60 recently opened accounts found that 41 had only one identity check, 22 above the value threshold lacked a source-of-funds rationale, and 17 were activated before sign-off. Front-office staff say the procedure arrived by email with no walkthrough, that the second check duplicates a system control they believe already runs, and that the sign-off step adds two days to onboarding while competitors open accounts same-day. The procedure was written by a compliance analyst who has since left. No one can explain why the second check is independent rather than automated.
1. Diagnose why this procedure failed, separating design defects from implementation defects.
Read the worked answer
The failure has both design and implementation roots, and they reinforce each other.
On design: the procedure was written without evidence that each step addresses a defined risk. The second independent identity check duplicates what staff believe is an existing system control, and no one can explain the rationale. If the duplication is real, the step is waste; if the system control is weaker than staff assume, the procedure failed to say so. Either way the requirement is unexplained, which invites circumvention. The source-of-funds rationale threshold is stated but the procedure apparently does not say what a sufficient rationale looks like, so staff have no standard to meet. The manager sign-off has no stated purpose, so it is experienced purely as delay.
On implementation: distribution by email with no walkthrough is not training. No one tested whether staff understood the steps. No monitoring was in place during the eight months, which is why a 60-account sample found failures at this scale. Front-office incentives reward speed and there is no counterbalancing signal that procedure adherence matters.
The two categories interact: a poorly justified procedure is easy to dismiss, and a dismissed procedure is never properly implemented. Fixing only the training will not work while the steps remain unexplained, and rewriting the procedure will not work without training and monitoring.
2. Propose a redesign of the procedure itself, and specify what evidence would show each step is justified.
Read the worked answer
Redesign should start from the risks the procedure is meant to address: impersonation and false identity, proceeds of crime, and unauthorised account activation. Each step should map to one of these.
For identity, determine what the system control actually does. If it performs a robust verification against an authoritative source, the manual second check should be removed or narrowed to exceptions such as failed matches, high-risk jurisdictions, or documents the system cannot validate. Evidence would be system testing results showing detection rates and false-negative rates, plus a sample of exceptions the manual check caught that the system missed.
For source of funds, replace the vague rationale with a defined standard: what documents or explanations are acceptable for different customer types and value bands, and who may approve an exception. Evidence would be a sample of files showing rationales that withstand scrutiny, and monitoring data on how often rationales are missing or thin.
For sign-off, state its purpose, for example confirming that identity and source-of-funds requirements are complete before funds can move. If the delay is the concern, consider parallel processing where activation is blocked until sign-off but preparation continues. Evidence would be the rate of pre-sign-off activations and the number of deficiencies caught at sign-off.
Document the rationale for each step in the procedure itself, so the next analyst inherits reasoning, not just rules.
3. Front-office leaders propose simply reissuing the procedure with a warning that non-compliance will be disciplined. Evaluate this proposal.
Read the worked answer
Reissuing with a threat addresses only the implementation side and only superficially. It assumes staff knew what to do and chose not to, but the evidence suggests they did not understand why the steps existed and doubted their value. A warning does not answer the question of whether the second check is duplicative, does not define an adequate source-of-funds rationale, and does not reduce the two-day delay that creates the pressure to skip sign-off.
It also risks a backlash. If staff comply grudgingly with steps they believe are pointless, they will comply minimally, and the quality of the underlying judgements will not improve. Discipline applied to a procedure nobody can justify looks arbitrary and erodes trust in compliance.
The better sequence is to fix the design first, explain the reasoning, train staff on the revised steps, and then attach consequences. Discipline is most credible when the requirement is clear, proportionate, and understood. The leaders' underlying concern, speed, is legitimate and should be met through process design rather than by tolerating non-compliance. A combined approach, redesign plus training plus monitoring plus clear accountability, is more likely to hold than a warning letter.