You might expect the information security policy to be the very first thing you write. After all, it sits at the top of all your ISMS documents. But in our roadmap, it only comes in Step 4, and there’s a good reason for that.
Clause 5.2 says the policy has to be appropriate to the purpose of the organization, and you only really know that purpose after Step 1, when you’ve looked at the context and set the scope. The policy includes your information security objectives, or the framework for setting them, and the gap analysis in Step 2 showed you where you stand. And most importantly, it’s top management’s policy, so you need their commitment from Step 3 first. Now it’s time, and everything that comes after Step 4 has to be consistent with it.
Step 4 is more than the policy, though. It covers three clauses: the objectives (6.2), the policy itself (5.2) and communication (7.4). And it starts with two processes that run in the background of everything else.
Start with two processes
The first is the policy and document management process. It’s the life cycle of every ISMS document: how it’s created, approved, published and reviewed. That’s clause 7.5 and Annex A control 5.1, and the information security policy is the first document that goes through it. If your organization already controls its documents, for example for ISO 9001, don’t build a second process next to it. Extend the one you have.
The second is the communication process for clause 7.4. Its first job is the communication plan, and the first thing on that plan is your new policy. More on that below.
1. Set the information security objectives
In the project plan, you determine the objectives before you write the policy, because the policy will either include them or describe how they’re set.
What is an information security objective? ISO/IEC 27003, the guidance standard for ISO 27001, keeps it simple: objectives for the confidentiality, integrity and availability of information. So “buy a new firewall” isn’t an objective. That’s a measure. An objective describes the result you want.
Clause 6.2 then asks for seven things. Your objectives are consistent with the policy, measurable where practicable, based on the applicable requirements and the results of the risk assessment and treatment, monitored, communicated, updated as needed, and available as documented information. So your objectives aren’t something you keep in your head. An auditor will ask to see them.
The most common objective there is: “protect the confidentiality, integrity and availability of our information.” It sounds right, and nobody would argue with it. But how would you ever know that you’ve achieved it?
Objective | Measurable? |
|---|---|
Protect the confidentiality, integrity and availability of company data | No |
Reduce the mean time to detect incidents to one hour | Yes |
Ensure 90% of employees complete the mandatory awareness training | Yes |
Obtain ISO/IEC 27001 certification | Yes |
The last one surprises people. But ISO/IEC 27003 explicitly mentions compliance with ISO 27001 as one way to express an objective, and it’s measurable: either you have the certificate, or you don’t. A simple check for the rest is SMART: specific, measurable, achievable, relevant and timely. ISO 27001 doesn’t mention SMART, but an objective that passes the test typically meets the “measurable” requirement without any discussion.
Clause 6.2 also wants objectives “at relevant functions and levels”. Typically, top management sets a few high-level objectives, and the departments derive their own from them.

Let’s say the high-level objective is that your critical business processes run again within 24 hours after a cyber attack. Then IT might commit to testing a full restore of the critical systems twice a year, HR to getting 95 percent of employees through phishing training, and purchasing to security requirements in every contract with a critical supplier. Each contributes to the objective above it.
A question that comes up quite often: clause 6.2 says the objectives take the results of the risk assessment into account. But you set them in Step 4, long before the risk assessment in Step 7. Is that the wrong order? Not really. ISO/IEC 27003 treats the risk results as an input to the ongoing review of the objectives. So you set your first objectives now, from your context, your interested parties and your gap analysis, and you review them after the risk treatment.
Plan how to reach them
The second half of clause 6.2 is the plan. For each objective, you determine what will be done, what resources it needs, who’s responsible, when it will be completed and how you’ll evaluate the results.

ISO 27001 doesn’t say where that plan goes. In practice, the activities and due dates typically don’t sit in a document at all. They sit in a tool: Jira in a larger organization, a task list in Excel in a smaller one. What matters is that you can show it. When an auditor asks how you’re going to reach an objective, you open the plan, and there it is: the activities, the people and the dates.
2. Write the information security policy
The clauses of ISO 27001 require exactly one policy. Not a password policy, not a backup policy, not an acceptable use policy. Just one: the information security policy.
It sits at the top of your documentation, like a pyramid. Below it come the topic-specific policies, for example on backups, access control or cryptography. Below those, the procedures that describe how something is done. And at the bottom, the records that show it was done.

Take backups. The policy says that critical business data is backed up regularly and can be restored. The topic-specific policy says daily. The procedure says when and by whom. And the backup logs show that it actually happened. In Step 4, you’re only at the top. The topic-specific policies come in Step 8, together with the controls they support. So keep password lengths and backup times out of the information security policy. Otherwise, top management has to approve a new version every time a technical detail changes.
Clause 5.2 names four things the policy has to contain:
It’s appropriate to the purpose of the organization. If you could swap your company name for another one, it isn’t appropriate yet.
It includes the information security objectives or the framework for setting them. A smaller organization might list its few objectives right in the policy. A larger one typically describes who sets them and how they cascade down.
A commitment to satisfy the applicable requirements: the legal, regulatory and contractual requirements from Step 1, and the standard itself.
A commitment to continual improvement of the ISMS.
What “top management establishes” really means
Clause 5.2 starts with three words that matter: top management establishes. Not the IT department, and not the ISMS manager. Someone else can draft it, but it’s top management’s policy. So the approval isn’t a signature at the very end.

It involves three things. Direction: the policy has to reflect top management’s direction, not just the draft of the ISMS manager. Distribution: top management decides which interested parties get the policy. And form: if the organization already has a policy template, the information security policy uses it. So the approval is the moment top management makes the policy its own. They discuss the draft, change what doesn’t reflect their direction, and then they approve it.
Then the policy has to reach people. It’s communicated to everyone within the scope, in a language and format they understand. People confirm that they’ve read it, which Annex A control 5.1 asks for. And it’s available to interested parties as appropriate: customers might ask for it. That’s why you write it so you can share it, with nothing confidential in it.
3. Plan your communication
Imagine a data breach. Who’s allowed to talk to the press? Who informs your customers? And who reports it to the authority, and within how many hours? If you’re figuring that out in the middle of the incident, it’s too late. Clause 7.4 is about deciding on your communication before you need it.

For each communication, you decide four things: what you communicate, when, with whom and how. Clause 7.4 tells you the questions, not the answers, and the answers can look completely different. A fully remote software company puts the policy on a wiki page and announces it in the team chat. A factory with shift work can’t reach the shop floor by email, so the team leaders brief each shift. A hospital makes it part of a short e-learning with a test, every year.
And here’s where a communication plan pays off, using the data breach from above:
With whom | When | Who |
|---|---|---|
Employees: report anything suspicious | Immediately, to the service desk hotline | Everyone |
Data protection authority | Within 72 hours (GDPR), if personal data is affected | The board, with the DPO |
Cyber authority | Early warning within 24 hours (NIS2) | ISMS manager |
Customers | As agreed in the contract | Account managers |
Media | Only if needed | CEO only |
Notice the last column. Clause 7.4 doesn’t ask who communicates. But ISO/IEC 27003 recommends deciding it, especially for cases like this one, and I’d say it’s one of the most useful columns in your plan.
At Barvello: “That’s not us.”
Barvello Coffee Systems is the fictional company of our ISO 27001 Lead Implementer course: a family-owned maker of professional coffee machines in Munich, with about 500 employees and its own cloud platform, Barvello Connect. In Step 3, the management board approved the project and made Nadia Berger responsible for the ISMS. The project charter already had the milestone for Step 4: publish the policy by 31 March. So Nadia worked backwards.
She started with the processes. Barvello has controlled its documents for twenty years under ISO 9001, and Anna Schulz, who runs that system, had told her in the gap analysis: “Please don’t build a second system next to mine.” So Nadia didn’t. The ISMS documents go into the same document center, they’re reviewed every January with the ISO 9001 documents, and Anna owns document control for both.
The communication plan turned out to be surprisingly concrete, because Nadia filled it with the requirements of the interested parties from Step 1. Ovena, the key customer, gets a security status report every quarter. The works council is consulted before any tool that can monitor employees goes live. A significant incident goes to the BSI, the German cybersecurity authority, with an early warning within 24 hours. And one line took her ten seconds to write: only the CEO speaks to the media. It might save Barvello a lot of trouble one day.
On 9 March, the management board set the objectives in the steering committee. Nadia had prepared them from four sources: the Barvello 2030 strategy, the project charter, the interested parties and the gap analysis. One came from Markus Lehner, the COO. In the business case, he had named two days as the limit for a production stop. Now it’s an objective: critical systems running again within 48 hours after a ransomware attack, measured by a restore test twice a year. The departments derived their own. IT will test a full restore of SAP by 30 June, the restore that was never tested in the gap analysis. The plan for each objective sits in the project plan: the insurer’s six conditions, for example, became six tasks with an owner and a due date before 30 June.
The first draft of the policy took Nadia ten days, because most of it already existed: the scope from Step 1, the roles from Step 3, the objectives and the communication plan from the last few weeks. For the objectives, she chose the framework option: the policy describes how objectives are set and points to the objectives document, so they can change without a new policy. And there are no password rules and no backup times in it.
On 22 March, the draft was on the agenda of the management board’s Monday meeting. Julia Keller, the CEO, read the purpose and said: “That’s not us.” She rewrote it herself, starting with Barvello’s customers and their coffee, because that’s how people at Barvello think about their work. The board also decided that the policy may be shared with customers on request, so it can’t contain anything confidential. On 29 March, Julia approved version 1.0.
On 31 March, she sent it to everyone by email. The team leaders read out a short version in the shift briefings, in German, and R&D and Barvello Inc. got it in English. Everyone confirms that they’ve read it in the learning platform. The works council had the draft a week earlier. And Ovena got the policy with its first quarterly status report in April. The milestone was reached on time, with a policy that the management board can explain in its own words.
Watch the full lecture above. It’s “Clause 5.2 Information Security Policy” from the course: what goes into the policy, what doesn’t, and what it means when top management approves it.
Your turn: Step 4 in the project plan
In the project plan of our ISO 27001 Toolkit, Step 4 is five tasks and one milestone:
ID | Task | Clause |
|---|---|---|
35 | Establish policy and document management process | 7.5, 4.4, A.5.1 |
36 | Establish communication process | 7.4, 4.4 |
37 | Determine information security objectives | 6.2 |
38 | Develop information security policy | 5.2 |
39 | Obtain approval for information security policy | 5.2 |
40 | Milestone: Publish information security policy | 5.2, A.5.1 |
At the end of Step 4, you have four deliverables, and the toolkit has a template for each. Two are required by ISO 27001: the information security objectives (REC-009) and the information security policy (POL-001). The other two are determined: the policy and document management process (PRO-003) and the communication process with its plan (PRO-004 and REC-008).

You’re done with Step 4 when your policy is approved by top management, published, and everyone in the scope knows about it.
Two of 19 required documents.
Here are the other 17.
Your policy and your objectives are two of the 19 documents and records ISO 27001 requires, and auditors expect 124 more. The checklist lists all of them, each mapped to its clause or Annex A control.

From the auditor’s chair
The first thing I check in a policy is whether it could belong to any other company. If you could swap the name and nothing else would change, it isn’t appropriate to the purpose of the organization yet.
The second is whether the people in the scope actually know it. A policy on the intranet that nobody has confirmed reading isn’t communicated. And I usually ask top management what’s in it. A policy they rewrote themselves sounds very different from one they only signed.
Frequently asked questions
How many policies does ISO 27001 require?
The clauses require exactly one: the information security policy. Annex A control 5.1 adds topic-specific policies, and which ones you need depends on your risks. They come in Step 8.
Do the objectives have to be in the policy?
No. The policy either includes the objectives or provides the framework for setting them. Many organizations choose the framework, so the objectives can change without a new policy.
Who has to approve the information security policy?
Top management. Someone else can draft it, but top management establishes it, which means it reflects their direction and they decide who receives it.
How often should the policy be reviewed?
At planned intervals and when significant changes occur, as Annex A control 5.1 asks. Many organizations review it once a year, together with their other management system documents.
Can we share the policy with customers?
Yes, and they’ll often ask for it. Clause 5.2 expects it to be available to interested parties as appropriate. So write it without anything confidential in it.
Step 4 in the course.
Watch the CEO rewrite it.
6 lectures (36 min) show you how to set measurable objectives, write a policy top management makes its own and plan your communication. Then you watch Nadia write Barvello’s policy, and the CEO make it hers. Your turn: 5 tasks in the project plan and the templates for the policy, the objectives and the communication plan.
Clause 5.2 Information Security Policy
05:57
Case Study: Barvello’s Information Security Policy
08:43
Template: Information Security Policy
POL-001
Template: Information Security Objectives
REC-009
With the policy published, you’re ready for Step 5: the asset inventory, where you find out which information and other assets your ISMS covers.


