Team & People
Security Policy Framework: Three Layers Staff Can Follow, With a Review Cycle and Exceptions
Advanced | 4 whiteboard pages · Task board: 5 columns (4 parent tasks, 12 subtasks)
Advanced: Assumes you know the technical terms and methods, and the steps are only outlined. Work through it on your own, guided by decision criteria and points to check.
Split your information security policy into three layers - policy, standards, and procedures - and set each document's owner, approver, review cycle, and exception flow. It suits stale or unread policies. Inventory documents by revision, owner, and readership; keep the policy short, standards company-wide, procedures in staff's words and screens. Map standards to NIST CSF 2.0 functions (Govern, Identify, Protect, Detect, Respond, Recover) to find gaps; verify them in NIST's latest docs. Rank items by feasibility and risk, cut or support unfollowable rules, and set who may accept which exceptions. Owners and an exception contact curb misfit copied rules, rules nobody keeps, quiet rule-breaking over tedious requests, and revisions that never reach staff. Examples change passwords only after a leak or concern and rank SMS below authenticator apps and passkeys. The 4-page whiteboard ends in a yearly PDCA review; the task board has 5 columns, 4 parent and 12 subtasks.
Once you have a free account, you can copy this template from the marketplace
Whiteboard preview
Read the whiteboard text as plain text
Security Policy Framework Board
How to use (4 pages): ① Set the goal and sort existing documents into 3 layers (sections①–⑤) → ② Write standards and procedures per area and check pitfalls (sections⑥–⑨) → ③ Narrow by feasibility and risk, then set exceptions and the request flow (sections⑩–⑪) → ④ Plan the yearly PDCA, rollout, and revision records (sections⑫–⑭)
1/4 Inventory current policy
① Goal and scope of the rework (start here)
- Goal of the rework e.g. Make the old policy something staff will read and follow. Be able to answer customers' security questionnaires by pointing to it
- Scope e.g. All company information, devices, and cloud use. Includes agreements with vendors
- Coordinator and timeline e.g. The IT manager coordinates, drafts the revision in 3 months, and gets approval at the management meeting
② Inventory of current documents (last revision, owner, read or not)
| Document | Last revised | Owner | Is it read? | |
|---|---|---|---|---|
| Document 1 | e.g. Information security policy statement | e.g. 7 years ago | e.g. General Affairs (the author left the company) | e.g. Read once at hiring |
| Document 2 | e.g. Device usage rules (40 pages) | e.g. 5 years ago | e.g. IT department | e.g. Almost no views in the intranet log |
| Document 3 | e.g. PC initial setup guide | e.g. Last year | e.g. IT staff member | e.g. Used every time a new device is set up |
| Document 4 | e.g. Remote work rules (internal notice) | e.g. 3 years ago | e.g. HR department | e.g. Only a notice, and nobody knows where the full text is |
③ Policy (why, and who is accountable)
- Role: a short document issued in management's name. Changes once every few years
- e.g. Put document 1 here. Keep only the purpose and the accountable person, cut to 2 pages or less
- Leave out: setting values and screen operations
④ Standards (what the whole company follows)
- Role: rules the whole company follows, one "do X" line per area
- e.g. Pull only the rules out of the 40 pages of document 2 and put them here
- Owned by the coordinator. Approved by management
⑤ Procedures (how it is done on the floor)
- Role: operations by department or job, written in the words and screens of the floor
- e.g. Put documents 3 and 4 here, as screenshots and numbered steps
- Owned by the department that uses it. Fix it as soon as a screen changes
2/4 Redraw it in three layers
⑥ Rewrite the rules in ④ as standards for 5 areas (NIST CSF 2.0 function and example procedure)
| Standard (what to follow) | CSF 2.0 function | Example procedure (floor screens and words) | |
|---|---|---|---|
| Accounts | e.g. Require multi-factor authentication for anything used from outside the company | e.g. Protect | e.g. Use an authenticator app or passkey. SMS is better than nothing but weak. Change a password after a leak or when in doubt |
| Devices and updates | e.g. Register work devices in a ledger and never let updates stall | e.g. Identify, Protect | e.g. Each month, check the device admin console for devices whose updates have stopped |
| Data protection | e.g. Keep copies of important data in a second place and test that they can be restored | e.g. Protect, Recover | e.g. Each quarter, restore one item from the backup console and record it |
| Vendors and cloud | e.g. Before signing with a vendor or cloud service that will hold our data, check what they will follow | e.g. Govern | e.g. Answer "Will data be stored there?" in the purchase request and attach a checklist if yes |
| Incident response | e.g. Whoever notices tells the contact point right away, and nobody blames the reporter | e.g. Detect, Respond, Recover | e.g. Do not click links in suspicious emails, and report by phone or chat to the contact point |
⑦ Pitfall: copying another company's policy that does not fit
- What happens: e.g. Rules for systems you do not use pile up, and staff stop reading the whole policy
- How to check: e.g. For each row of ⑥, see whether that system exists in the current ledger or admin console
- How to prevent: e.g. Delete rows for systems not in the ledger and rewrite them in your own words
- Rows in ⑥ that apply: e.g. The vendor and cloud checklist is still another company's
⑧ Pitfall: rules that were set but cannot be followed
- What happens: e.g. Rules nobody follows stay on the books, and other rules start to be taken lightly
- How to check: e.g. See the rate of compliance in admin console settings and the exception register
- How to prevent: e.g. Put it on the left side of ⑩, then add support, cut it, or make it a time-limited exception
- Rows in ⑥ that apply: e.g. "Devices and updates" with a PC running old business software
⑨ Pitfall: exception requests so tedious that rules are broken silently
- What happens: e.g. People drop a rule without applying, and nobody knows where the exceptions are
- How to check: e.g. Compare the count in the exception register with the number of settings that break the rules in the admin console
- How to prevent: e.g. Make the request in ⑪ one line of reasoning plus a deadline
- Rows in ⑥ that apply: e.g. Shared accounts under "Accounts"
3/4 Narrow to what is doable
⑩ Rank the standards of ⑥ by ease of following and risk (also use the results of ⑦–⑨)
Vertical axis: Risk if not followed (low at the bottom, high at the top) / Horizontal axis: Ease of following (low on the left, high on the right)
- Add support to make it followable e.g. MFA for everyone. Prepare how to hand out authenticator apps and a guide first, and allow time-limited exceptions only during the transition
- Keep as required e.g. Disable a leaver's account the same day. Write it as one line in the standards and allow no exceptions
- Cut or lower to recommended e.g. Items that bind the fine settings of personal devices. Lower to recommended, and give guidance in procedures if needed
- Leave to procedures e.g. The time before screen lock. In the standards write only "lock the screen automatically" and set the value in the ⑤ procedures
⑪ Bar for accepting exceptions and the request flow (for items placed on the left of ⑩)
| Severity low (one department) | Severity medium (several departments, partners) | Severity high (customer data, business stop) | |
|---|---|---|---|
| Request | e.g. One line of reason on the request form | e.g. Reason and a substitute control | e.g. Reason, substitute control, and scope of impact |
| Approval | e.g. Department head | e.g. Information security officer | e.g. Management (items that cannot be accepted go to the top right of ⑩) |
| Deadline | e.g. 1 year at most | e.g. 6 months at most | e.g. Review within 3 months at most |
| Record | e.g. One line in the exception register | e.g. Enter in the register and check each quarter | e.g. Enter in the register and report to the management meeting |
4/4 Roll out and review yearly
⑫ Review the policy narrowed in ⑩ and ⑪ every year with PDCA
- Plan Draft e.g. In Q1 each year, gather the exception register, incident records, and partner questions, and draft revisions
- Do Roll out e.g. Get approval, fix procedures and screen explanations, and communicate as scheduled in ⑬
- Check Verify e.g. After 6 months, check compliance through admin console settings, the exception register, and inquiries
- Act Adjust e.g. Send items nobody follows back to ⑩ and choose among cut, add support, or make an exception
⑬ Rollout schedule (to whom, when, how, and how you confirm)
| When | How to tell them | How to confirm it landed | |
|---|---|---|---|
| Executives | e.g. At the approval meeting | e.g. Explain what changed in 5 minutes | e.g. Record questions and decisions in the minutes |
| Managers | e.g. The week after approval | e.g. Explain the changed procedures at department meetings | e.g. Log in the register when each department was briefed |
| All staff | e.g. 2 weeks after approval | e.g. Post the changes in 3 lines on the intranet | e.g. Confirm they read it and collect answers to 3 short questions |
| Vendors and temp staff | e.g. By the next contract renewal | e.g. The person in charge hands them a document with the changes | e.g. Record the reply confirming receipt |
⑭ How to record revisions, and the pitfall of not communicating them
- Revision record (one line per revision) e.g. Version 3.0 / Year 2 Q2 / Added the exception rules / Approved at the management meeting / Who was told and when is in the table of ⑬ (all 4 rows done)
- Pitfall: revisions that are never communicated What happens: e.g. Staff follow the old procedures and nobody knows the new rules How to check: e.g. See the results of the checks in ⑬ and whether inquiries mention the old version
- How to prevent e.g. Add a "who was told and when" field to the revision record, and do not publish the new version while it is empty
Task board contents
Inventorying 16 items
Audit what exists
On page 1, set the goal and scope, then list the existing documents in a table and sort each into policy, standards, or procedures. Done when you can see which documents to keep and which to rewrite. Press Ctrl+F (⌘F on Mac) and enter "3/4" to jump to page 3.
List the documents
Open the shared folder or intranet portal, search for "policy", "security", and "procedure", and copy each document's name, last revision, and owner into the table in section ②. Done when 4 rows are filled in about 15 minutes. Leave the readership column empty.
Set goal and scope
In the 3 notes in section ①, write the goal of the rework, the scope, and who coordinates it and by when. Done when you know who will approve it, such as the executive team.
Sort into 3 layers
For each document in section ②, write on a note which of sections ③–⑤ it belongs to. If one document mixes policy and procedure, split it and write it under both sections.
Redraw in 3 layers
On page 2, rewrite the rules sorted into ④ as standards for 5 areas, and add example procedures. Done when all 3 layers have content and the pitfall check is finished.
Write standards
In the 5 rows of section ⑥, write one standard per row in the form "do X", and add an example procedure in the words and screens staff use. Done when all 5 rows are filled.
Map to CSF 2.0
In column 2 of section ⑥, write which NIST CSF 2.0 function each standard falls under. If a function matches no row, add a note below section ⑥ as a thin area. Check function names and contents against the latest NIST material.
Check the pitfalls
For each of the 3 pitfalls in sections ⑦–⑨, list the rows in section ⑥ that apply and decide one step to prevent it. If no row applies, write "none" and you are done.
Narrow to doable
On page 3, rank the standards by how feasible they are to follow and by risk, then decide what to do with rules nobody can follow, the bar for accepting exceptions, and the request flow. This is the core work of balancing the weight of the policy against the effort on the floor.
Rank ease vs. risk
Place each row of section ⑥ in one of the 4 areas of section ⑩. Cut items at the bottom left or lower them to recommended, and write supports (tools, procedures, a transition period) for items at the top left.
Define exceptions
In the approval row of section ⑪, write who can accept each severity and get management sign-off. Done when you can say in one sentence who accepts up to which severity.
Try a request
Following the request, deadline, and record rows of section ⑪, write one existing exception (e.g. a PC running old business software that cannot be updated) as a request. Rewrite the points where you hesitated into section ⑪.
Roll out, review
On page 4, decide the yearly review flow, the rollout schedule, and how revisions are recorded. Do not treat the policy as finished once written; leave it ready to review on the same page next year.
Plan the rollout
In the table of section ⑬, write for each audience when and how you tell them and how you check it landed. Done when all 4 rows have timing, method, and check.
Log the revision
Write this revision as one line in the form of the first note in section ⑭. If the fields for who was told and when are empty, do not publish yet.
Plan yearly review
Decide the timing of Plan in section ⑫ and the records to review in Check (admin console settings, exception register, inquiries), then add them to the company calendar.