A charter operator does not wait for the client to ask whether the aircraft is ready. The trip sheet, crew duty records, catering and handling arrangements are checked before wheels-up, and the paperwork is organised so a question on the ramp gets an answer in seconds. ComplianceOps applies the same instinct to rules and audits in technology businesses: stay ready every day instead of cramming when an inspector appears. Use the checklist below as a self-audit. It is general operational information, not legal or regulatory advice.
Before you start: what a ComplianceOps audit actually tests
The self-audit is not asking “do we have policies?” It asks whether each promised check is happening, whether you can show it, and whether the proof is recent. The Ops Atlas guide to ComplianceOps defines the aim as being able to serve an audit from records that already exist. Keep that benchmark in mind as you go through the items.
Keep your newest records open next to you as you go, since a self-audit done from memory tells you only what you remember. Score every line as pass, fail or untested. Untested is a legitimate answer and must be counted separately, never rounded up to pass.
Section A: the list of obligations
- 1. One register exists. Every rule, contract duty and framework requirement you answer to is written in a single list.
- 2. Each line has a named owner. A department is not an owner. A person or role is.
- 3. The register was reviewed within the last quarter. New customers and new partners add obligations, and lists drift.
Section B: the controls
- 4. Each obligation maps to at least one control. If a requirement has no control, mark it as a gap and do not hide it.
- 5. Controls are written as testable statements. “Crew records are reviewed every month” can be tested; “we take duty limits seriously” cannot.
- 6. Controls shared across frameworks are recorded once. Separate teams re-asking operations the same question for each rulebook is a classic waste the Atlas warns about.
- 7. Mapping notes record intervals. Where several rules ask for periodic review with different timings, the note shows which interval each rule assumes.
Section C: the evidence
- 8. Each record shows source, date, method and test. Anything missing one of these is weak.
- 9. A freshness window is set. Define how old a record may be before it stops counting.
- 10. Stale records are scored as missing. Audit-day optimism does not move a stale item to pass.
- 11. Records are superseded, not edited. If a number changes, add a new record and keep the old one visible. Altering the original is a serious red flag.
- 12. Collection began before the audit period did. Starting a reporting window before your collectors were connected leaves holes you must then explain.
Section D: exceptions and gaps
- 13. Every known failure is covered by a live exception. The exception names the exact resources affected.
- 14. Each exception has an end date. Once expired, it is a failure again.
- 15. No exception has been renewed just to turn a report green. A compensating safeguard must accompany any extension.
- 16. Missed records follow the gap runbook. Never back-date. Rebuild the record, label it as retrospective with the true date of collection, and tell the reviewer.
Section E: suppliers and third parties
- 17. Supplier assurance reports are read, not just filed. Look for the duties that the report expects you to perform yourself.
- 18. Letters bridging a gap in a supplier’s reporting period are treated for what they are. Per the Atlas, such letters come from the supplier’s management and are not validated by the auditor.
- 19. Questionnaire answers cite references. Each answer should point to a control and an evidence record.
Section F: reporting honestly
- 20. Reports show the untested count beside the pass rate.
- 21. Coverage and proof are kept separate. A control can be mapped and still unproven.
- 22. Nobody calls the result “certified” when it is an attestation. The words matter.
Scoring your results
Count the passes, fails and untested items separately. If more than half your checks rely on manual collection, that is a signal to automate the easiest ones first; the Atlas treats an automated majority as a marker of a more mature programme. A sensible rhythm for a small operator is to put an owner and a register in place in the first month, get collectors and tests running for the ten most important controls in the second, and only then open a formal reporting window once everything in it is fresh.
Remember the limits. A clean checklist shows that agreed controls operated when sampled. It does not show they were the right controls. A well-known retail breach in 2013 followed a recent payment-card certification, which is why the Atlas insists that compliance and security are different.
Where to read further on operations
If your own business runs ground transport alongside flights, our note on planning road legs across Canadian highways covers the practical side of getting around. And for a lighter weekend angle, see the post about getting out on the open road with a rental.
Run this audit once, fix the failures, and put the next review in the calendar. Nobody expects a spotless first pass. What matters is that the next run is better than this one, and that every red or amber line carries an owner and a date for being closed. Share the results with the people who own each section, because a score only changes behaviour when the person responsible can see it.

