Governance policy and implementation evidence register
Keep five questions visible when recording a governance policy: which version was approved, where it applies, who must carry out the linked control, what records show that work, and what a separate review establishes about the intended result. One green policy status cannot answer all five.
The eight-page pack contains instructions, five reusable blank sheets and a two-page fictional worked example. The DOCX is editable and the matching PDF is for printing. Both are manually maintained records. They contain no formulas, automated validation, reminders, scoring, signatures or transmissions.
Download the editable governance policy and implementation evidence register (DOCX).
Download the print governance policy and implementation evidence register (PDF).
Give the policy and its applicability separate records
Begin with one policy family and a stable version ID. Record its approval authority, approval evidence and effective dates as stated in the source. Mark an unapproved draft as a draft. A file uploaded recently may contain an older policy; its upload date does not establish the date from which it applied.
The applicability sheet identifies the entities, sites, committees, roles or activities covered by that version, together with dates and the basis for each inclusion or exclusion. An unresolved boundary stays unresolved. A group policy heading does not, by itself, show that every subsidiary adopted it or that a newly acquired operation was covered throughout the period.
Keep the exact commitment and section locator beside the record. Split commitments when their operating controls, owners or relevant populations differ. Store the source in its authorised location and record a usable reference. A register should not become a public collection of confidential minutes, personnel records or legal advice.
Connect each commitment to work that can leave a record
On the control sheet, describe the required task, responsible role, trigger or frequency, expected population and record to retain. Link the procedure or form version that tells the owner how to do it. Name the reviewer separately where the actual process requires one. Recording an assignment establishes who was assigned; it does not establish that the task happened.
A control can serve more than one commitment. Keep one control ID and add the policy-version and commitment links. Avoid duplicate control rows that imply different owners or schedules for the same work. If a policy changes, review these links and record what actually changed in the procedure, scope or evidence expectation.
OECD’s 2023 governance principles include board attention to implementation and the effectiveness of governance practices. This is narrow context for keeping evidence questions separate; the principles do not prescribe this register.
Record operation before judging the result
For each expected occurrence, identify the control, location or group, date or period, evidence version and exact locator. Say whether the relevant record was examined, remains unavailable or contains an exception. Preserve the actual event date separately from the date the document was received. A populated form can still omit a required field or show a task performed late.
The effectiveness sheet asks a different question: did the control achieve its stated purpose under the reviewed conditions? Define the test, population, selection method and limits before recording a conclusion. Use a qualified reviewer when the question requires expertise. A small, deliberately selected sample does not become a representative result simply because every selected record was examined.
Record the observed result and the permitted conclusion in words. Use “not assessed” when no effectiveness review occurred. Use “inconclusive” when the available evidence cannot answer the question. A completed implementation task can coexist with either state. Do not collapse missing evidence, an established missed task and an unsuccessful test into one unexplained score.
Work through an ordinary governance decision record
All organisations, people, policies, records and dates in this example are invented. At Vale Workshop, policy P4 v1 was approved on 18 August 2026 and took effect on 1 September. It covers the East and West project steering committees. It does not cover the North informal working group. The case examines September records only.
Commitment K1 requires each covered meeting’s decision record to state the decision, reason and action owner, with a chair review before the record is finalised. C1 assigns preparation to each committee secretary. C2 assigns the completeness review to its chair. The fictional calendar identifies four covered meetings: E1, E2, W1 and W2. A blank form and the North group’s note are outside that four-record population.
The packet includes a completed form for each of the four covered meetings. E1, E2 and W1 have a recorded chair review. For W2, no chair-review record is located; the owner has not established whether the review occurred. W1 has a chair mark but its reason field says only “as discussed”. This is an identified content exception, even though a signature is present.
The records therefore support four prepared forms and three recorded chair reviews within the defined population. They do not establish four satisfactory completeness reviews. W2 has a missing-evidence question, and W1 has an observed rationale gap. Keep both open without inferring that anyone deliberately bypassed the process.
The agreed effectiveness exercise asks whether a reviewer who did not attend can reconstruct the decision’s reason and identify the action owner from the retained record and its cited pack. Reviewer Mina deliberately selects E1 and W1, one from each committee. This is a two-record diagnostic exercise, not a statistically representative sample.
For E1, the cited pack explains the access disruption that led to rescheduling a maintenance review, and the decision record names the facilities lead as owner. Mina reconstructs both from the supplied records. For W1, the pack lists an agenda but gives no reason for changing the review sequence. The action owner is named, but the reason cannot be reconstructed. The test supports a narrow positive result for E1 and an unresolved rationale gap for W1; it does not establish policy-wide effectiveness.
Keep the decision and the next check traceable
Decision D1 records the policy’s approval and East/West applicability separately from the control findings. It keeps W2’s chair-review evidence request and W1’s rationale gap open. The West secretary owns both follow-ups. The agreed date is 12 October 2026, 16:00 IST. The original W1 record and the chair mark are retained.
The proposed W1 correction is to add a dated explanatory note supported by existing records and reviewed by the chair. It must not be presented as the original contemporaneous reason unless that is established. The proposed process change adds a specific rationale check to the chair’s checklist. Neither proposal has been completed at the case cut-off. Even completing both would not prove that later meeting records are usable; the planned follow-up test remains separate.
The decision-history sheet gives each new conclusion its own event ID, source versions, reviewer and date. Use it when evidence arrives, a policy is superseded or a conclusion changes. Identify dependent summaries or disclosures that need rechecking. Updating a current-status cell without preserving the earlier decision loses the reason a reviewer relied on the previous evidence.
Use the register with the existing review process
The policy-version guide covers version and rollout mismatches. This download supplies the reusable records needed to carry a commitment through applicability, assigned controls, operating evidence and a bounded effectiveness review.
Use the source document and approval register for the controlled source inventory. Use the corrective-action closure guide when a gap needs a process change, and the management sign-off checklist for a separate release decision. This register records evidence; it does not authorise publication or clear a governance requirement.
Source checked 15 October 2026: OECD (2023), G20/OECD Principles of Corporate Governance 2023, V.D.1 and V.D.3, printed page 36. © OECD 2023.
Version 1.0. Original Nitium Advisory forms, diagram and fictional example. This optional working aid does not determine legal applicability, establish compliance, confer certification or provide independent assurance. Adapt it to the organisation’s approved process. No source organisation endorses it. Remove the fictional pages before live use.
Comments