PLAYBOOK

Cross-border dispute resolution

Dispute taxonomy, evidence packs, and recovery options by jurisdiction.

Cross-border dispute resolution

This playbook turns a recurring trade-operations workflow into a sequence of named steps with measurable exits. Use it as a working document — copy it into your ops handbook, adjust the steps to your corridor, and run it.

When to use this playbook

Use this playbook when you're standardising a workflow that's currently living in someone's head, when a new hire needs to ramp on the procedure, or when the existing process is producing inconsistent outcomes across operators.

Prerequisites

Access to the relevant systems (ERP, CRM, payment portal, customs platform)
A named owner accountable for outcomes
Baseline metrics on the current state — without these, "did the playbook help?" is unanswerable
Templated artefacts: PO, invoice, contract, dunning letters, KYB pack

The procedure

Step 1 — Define the trigger

State the precise event that kicks off the workflow. Vagueness here cascades into every downstream step.

Step 2 — Map the data flow

For every step, name the inputs (where the data comes from) and outputs (where it lands). Most failed playbooks fail at this layer, not at the human-judgement layer.

Step 3 — Apply the controls

Insert the verification checks: payment-instruction confirmation, sanctions screening hits, KYB decisions, document compliance vs UCP 600. The right answer is rarely "skip it because we know the buyer."

Step 4 — Decision rule

State, in advance, how the workflow branches. "Decline if KYB returns adverse media + sanctions hit + unverified UBO" is a decision rule. "Use judgement" is not.

Step 5 — Execute and log

Every meaningful action ends up in the audit log with a UETR, invoice ID, or case ID. Audit-traceability is the difference between defensible operations and reconstruction-from-memory.

Step 6 — Review and tune

Weekly: pull the metrics, compare to the baseline, identify the bottleneck. Monthly: tune the decision rules, retrain the AI components if used.

Common pitfalls

Steps that depend on a single person's tribal knowledge
Decision rules that lean on "judgement" without a written rubric
No instrumentation, so no feedback loop
Templates that drift over time without versioning

AI / automation hooks

Most steps in a modern operator playbook are now augmentable: KYB decisions, dunning cadence selection, HS classification, cash application matching, and exception triage. See Taxonomy of AI agents in trade for the patterns.