# Policy vs Procedure: Differences, Standards and a Security Example

Source: https://truvocyber.com/blog/policy-vs-procedure
Published: 2026-11-21
Topic cluster: Cybersecurity

> Policy vs procedure: a policy states what and why, a procedure states how. Comparison table with standards and guidelines, plus an access review example.

In policy vs procedure terms, a policy states what an organization commits to and why. Leadership approves the policy, and the text changes rarely. A procedure states how a team meets that commitment, step by step. The team doing the work owns the procedure and updates the steps whenever a system or tool changes.

At a glance

- **Policy**: What and why. Approved by leadership. Reviewed about once a year. Acknowledged by personnel
- **Standard**: The mandatory specifics behind a policy, such as MFA methods, TLS versions or hardening baselines
- **Procedure**: How, step by step. Owned by the team that does the work. Updated whenever the process or system changes
- **Guideline**: Recommended practice, with room for judgment. Optional
- **Quick test**: A document that names hostnames, IP addresses or dashboard steps is a procedure
- **Framework fit**: SOC 2 CC5.3, ISO/IEC 27001:2022 Clause 5.2, A.5.1 and A.5.37

## What is the difference between a policy and a procedure?

A policy is a short statement of commitment: what the organization will do, who is accountable and why. A procedure is the set of steps a named team follows to meet that commitment, with the tools, systems and records involved. The two documents differ on five points:

- **Purpose:** the policy sets the expectation. The procedure describes the execution.
- **Approval:** leadership or a security committee approves the policy. The process owner approves the procedure.
- **Audience:** everyone in scope reads and acknowledges the policy. The people who perform the task use the procedure.
- **Change rate:** a policy changes when the business, the risk or a law changes. A procedure changes when a tool, a system or a step changes.
- **Detail:** the policy names roles and outcomes. The procedure names systems, screens, commands and records.

Large organizations enforce the split because a policy change goes through executive approval and formal review cycles, while procedures need frequent updates without that overhead. In small companies the reason is maintainability: a policy full of system detail needs a formal update every time the infrastructure changes.

## Policy vs procedure vs standard vs guideline: how do the four security documents compare?

Policies, standards, procedures and guidelines form a hierarchy of security documents, from the broad commitment down to the specific step. Access control is the example in every row:
DocumentQuestion answeredMandatory?Approved or owned byTypical change rateAccess control examplePolicyWhat and whyYes, for everyone in scopeLeadership or security committeeAnnual review, or after a significant changeUser access to in-scope systems is reviewed at least quarterlyStandardWhich exact settingsYesSecurity ownerWhen technology or threats changeMFA is required on the identity provider for every user accountProcedureHow, step by stepYes, for the people doing the taskProcess owner, such as IT or engineeringWhenever a tool, system or step changesExport users from the identity provider, compare against the HR roster, open a review task for each system ownerGuidelineWhat the security team recommendsNoSecurity teamAs neededGrant access through groups instead of to individual accounts

### What is a security standard, compared with a security policy?

A security standard sets the mandatory, measurable specifics that make a policy concrete. Examples include:

- Password length and MFA methods
- Approved encryption algorithms and the minimum TLS version
- Hardening baselines such as the [CIS Benchmarks](https://www.cisecurity.org/cis-benchmarks)
- Logging fields and log retention periods

A security policy says data in transit is encrypted. The standard names the minimum TLS version. Standards change more often than policies and less often than procedures. Small companies often fold standards into a short table in the policy or into the program manual.

### What is a guideline, compared with a procedure?

A guideline is advice the team is encouraged to follow, with room for judgment. A procedure is mandatory for the people who perform the task. Keep guidelines out of the policy text. Auditors test a policy as written, and a recommendation phrased inside a policy can be read as a commitment.

## What does a policy vs procedure example look like for access reviews?

An access control policy clause states the commitment in one or two sentences, and the access review procedure turns that clause into steps, systems and records. An example clause in Truvo house style:

**Example policy clause: access control**

**Access Reviews.** User access to in-scope systems is reviewed at least quarterly by the system owner. Access that is no longer required is removed, and each review is recorded.

The access review procedure for the same clause names the tools and the order of work:

1. Open the quarterly access review ticket on the first business day of the quarter and assign the ticket to the IT lead.
2. Export the user list from the identity provider and from each in-scope system with local accounts, for example the cloud console, the production database and the source code platform.
3. Compare each export against the current HR roster of employees and contractors.
4. Send each system owner a review task listing every account and role.
5. The system owner marks each account keep, modify or remove.
6. IT makes the changes and attaches the exports, decisions and sign-off to the quarterly ticket.

Every detail that changes when a tool changes is in the procedure: the identity provider, the export steps, the ticket queue. When the company moves to a new identity provider, the team rewrites step 2 and the policy clause stays the same. The [SOC 2 user access review playbook](/blog/soc2-personnel-onboarding-access-reviews-saas-cto) covers the full review and onboarding process.

Key insight

**A document that names hostnames, IP addresses or dashboard steps is a procedure.**

System names, network addresses and click paths change with the infrastructure. Keep them in procedures, and keep the policy to commitments, roles and outcomes.

## Why do security policies end up full of procedure detail?

Security policies collect procedure detail when a team writes the first policy set in one pass and puts everything known about a process into the same document. At one SaaS company of 15 to 20 people writing policies for SOC 2, the disaster recovery policy held data center hostnames, IP addresses, dashboard instructions and failover steps next to the policy commitments. Any infrastructure change would have triggered a full policy review.

The fix for that disaster recovery policy took three steps:

- **Apply the hostname test.** Content that mentions hostnames, IP addresses or specific dashboard navigation moved to a disaster recovery procedure.
- **Use one GRC platform module for both documents.** The policy and the procedure lived in the same module with different acknowledgment settings: personnel acknowledge the policy annually, and the team updates the procedure as needed with no acknowledgment cycle.
- **Keep short steps where they are.** A few lines of procedure inside a policy are fine for a small team. Anything growing into a multi-page technical document moves out.

Policy templates add a second source of misplaced detail. Templates from compliance platforms often blend requirements from several frameworks and commit the company to tiers, SLAs and tooling the company does not operate.

## How does procedure detail in a policy cause audit findings?

Procedure detail in a policy causes audit findings because an auditor tests the policy as written. Every step, tool name, cadence and absolute statement inside a policy becomes a commitment the team has to show evidence for, and any gap between the text and the work becomes an exception, even when the security outcome is sound. Four examples:

- **Stale system detail.** A policy naming a tool the team has since replaced describes a process that no longer exists. The [ISO 27001 policy evidence gap post](/blog/iso-27001-policy-evidence-gap) covers the wrong-tool mismatch in detail.
- **Absolute statements.** At one identity verification SaaS company with under 10 people, the encryption policy first said all data in transit is encrypted. Exceptions existed for internal DNS, ACME certificate requests and HTTP redirects, so the absolute statement would have turned normal operations into policy violations. Changing the clause to encryption in transit where feasible removed those potential findings.
- **Commitments from the wrong framework.** The same company's privacy policy came from a template that blended GDPR and SOC 2, and committed to data portability. The company deletes all data within seven days. The team traced every requirement back to the framework that demands the requirement, then removed or qualified anything outside scope, and documented where the architecture already satisfies a provision.
- **Inherited cadences.** A vulnerability management template set tiered scanning schedules. For a self-hosted environment of three servers, the team collapsed the tiers into one weekly cadence for internal host scans and external network scans across production and staging, a cadence the team could keep.

Watch out

**Auditors test the policy text, line by line.**

A weekly cadence in a policy calls for a scan history with a scan every week. A hostname in a policy calls for evidence from that host. Write only commitments the team meets today, and move the method into procedures.

For turning existing policies into recurring work with owners and triggers, see the [guide to operationalizing security policies](/blog/operationalizing-security-policies). For the evidence side, the [SOC 2 evidence gap post](/blog/soc2-security-documentation) covers how auditors rank logs, screenshots and written descriptions.

## Policies and procedures for SOC 2: what order should they be written in?

My approach to policies and procedures for SOC 2 is manual-first: agree on how the work happens, record that in the Security Program Manual, and derive the policy last. Each security capability becomes four linked artifacts, built in this order:

1. **Security Program Manual section.** The concrete operational reality agreed with the team: which tool, what cadence, who triages, and what happens when a deadline cannot be met. The manual holds the procedures, tools and schedules.
2. **Policy.** Derived from the manual. Short, stating commitments, roles and deadlines.
3. **Control statements.** Each policy commitment rewritten as a testable control.
4. **Test definitions and evidence.** For each control, the test and the exact evidence artifacts, named before the audit.

Teams that start from a policy template and work downward tend to get stuck. Policy language is abstract, and people struggle to agree on words that do not yet describe anything. Once the team agrees on the activity, the policy is quick to write.

In one SOC 2 build for a company on self-hosted bare metal, the team assembled vulnerability and patch management as one artifact set in one working session. The evidence was named in advance:

- A screenshot of the scan tool configuration showing scope and frequency
- Tickets for Critical and High remediation
- A list of Critical vulnerabilities at the close of each audit window
- Ticketed records of weekly patching for host infrastructure

The policy gained an evidence and reporting subsection, so the retention commitment is in the document the auditor reads: records kept for the observation period plus three months, because a Type 2 audit finishes after the observation window closes. After every working session, a consistency review checked that an update in one artifact did not contradict another, because auditors notice drift between documents that describe one program.

The [vulnerability management policy template](/blog/vulnerability-management-policy-template) and the [incident response policy template](/blog/incident-response-policy-template) follow the same split: short policy clauses, with tools, schedules and steps in the Security Program Manual.

Key insight

**Write the Security Program Manual first and the policy last.**

A policy derived from agreed, working procedures holds only commitments the team already meets, so the first audit sample matches the text.

## What SOC 2 and ISO 27001 evidence comes from a clean policy and procedure split?

A clean split between policies and procedures produces the documents auditors ask for under SOC 2 and ISO 27001 as a byproduct of the work:
FrameworkRequirementEvidence from the policy and procedure split[SOC 2 CC5.3](/blog/soc-2-cc5-3-policies-procedures)The entity deploys control activities through policies that establish what is expected and in procedures that put policies into action ([AICPA Trust Services Criteria](https://www.aicpa.org/resources/download/trust-services-criteria))Approved policy set with owners and versions, procedures linked to each policy, records from each procedure runISO/IEC 27001:2022 Clause 5.2Top management establishes an information security policy ([ISO/IEC 27001:2022](https://www.iso.org/standard/27001))Information security policy with a named top management approver and dateISO/IEC 27001:2022 Annex A 5.1Information security policy and topic-specific policies are defined, approved by management, published, communicated, acknowledged and reviewedTopic-specific policies, acknowledgment records, review historyISO/IEC 27001:2022 Annex A 5.37Operating procedures are documented and made available to the personnel who need themProcedures in the Security Program Manual, with access for the teams that use them

## Where does policy and procedure work fit in an effective security program?

Policy and procedure work belongs inside each security capability, one of the roughly 19 building blocks of an effective security program. Truvo Cyber starts with a [security capability assessment](/services/security-capability-assessment) to score each capability, then writes the Security Program Manual, the policies, the control statements and the evidence definitions as one linked set. For mapping policies to controls and automating evidence in a GRC platform, see [GRC engineering services](/services/grc-engineering-services).

A policy says what the organization promises. A procedure shows how the team keeps the promise. Keep the two apart, and both stay true.

### Keep policies short and procedures current

Truvo builds an effective security program with a program manual, short policies and the evidence auditors sample.[Book Your Strategy Call](/contact-us)

## Frequently asked questions

### What is the difference between a policy and a procedure?

A policy states what an organization commits to and why, is approved by leadership and changes rarely. A procedure states how a team meets that commitment, step by step, with the tools and records involved. The team that does the work owns the procedure and updates the steps whenever a system or tool changes.

### What is the difference between a policy, a standard and a procedure?

A policy sets the commitment, such as encrypting data in transit. A standard sets the mandatory specifics behind the policy, such as the minimum TLS version or required MFA methods. A procedure gives the steps to apply the standard on real systems. Guidelines sit beside all three as optional, recommended practice.

### Does SOC 2 require separate policies and procedures?

SOC 2 CC5.3 expects control activities deployed through policies that establish what is expected and procedures that put the policies into action. The organization chooses the format. Many small companies keep short procedural steps inside a policy and move longer technical instructions, such as runbooks with hostnames, into separate procedures or a program manual.

### Who approves security policies and procedures?

Leadership approves security policies, often through a security committee, and ISO 27001 Clause 5.2 assigns the information security policy to top management. The process owner, such as the IT lead or the engineering manager, approves procedures. Splitting approval this way lets teams update procedures quickly without reopening executive sign-off.

### How often should policies and procedures be reviewed?

Common practice is a policy review at least once a year and after any significant change. ISO 27001 Annex A 5.1 calls for review at planned intervals and when significant changes occur. Procedures get updated whenever a tool, system or step changes, with a periodic check that each procedure still matches how the team works.

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "What is the difference between a policy and a procedure?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "A policy states what an organization commits to and why, is approved by leadership and changes rarely. A procedure states how a team meets that commitment, step by step, with the tools and records involved. The team that does the work owns the procedure and updates the steps whenever a system or tool changes."
      }
    },
    {
      "@type": "Question",
      "name": "What is the difference between a policy, a standard and a procedure?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "A policy sets the commitment, such as encrypting data in transit. A standard sets the mandatory specifics behind the policy, such as the minimum TLS version or required MFA methods. A procedure gives the steps to apply the standard on real systems. Guidelines sit beside all three as optional, recommended practice."
      }
    },
    {
      "@type": "Question",
      "name": "Does SOC 2 require separate policies and procedures?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "SOC 2 CC5.3 expects control activities deployed through policies that establish what is expected and procedures that put the policies into action. The organization chooses the format. Many small companies keep short procedural steps inside a policy and move longer technical instructions, such as runbooks with hostnames, into separate procedures or a program manual."
      }
    },
    {
      "@type": "Question",
      "name": "Who approves security policies and procedures?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Leadership approves security policies, often through a security committee, and ISO 27001 Clause 5.2 assigns the information security policy to top management. The process owner, such as the IT lead or the engineering manager, approves procedures. Splitting approval this way lets teams update procedures quickly without reopening executive sign-off."
      }
    },
    {
      "@type": "Question",
      "name": "How often should policies and procedures be reviewed?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Common practice is a policy review at least once a year and after any significant change. ISO 27001 Annex A 5.1 calls for review at planned intervals and when significant changes occur. Procedures get updated whenever a tool, system or step changes, with a periodic check that each procedure still matches how the team works."
      }
    }
  ]
}
