# Runbook vs Playbook: Differences, Examples and a Template

> Runbook vs playbook in plain words: what each covers, how SRE and SOC teams use the terms, a worked runbook example, a template and how to keep them current.

Canonical: https://origindatapartners.com/guides/runbook-vs-playbook
By: Origin Data Partners
Status: Published
Published: 2026-11-17
Updated: 2026-11-17
Sources checked: 2026-11-17

## Quick answer

A runbook is a step-by-step guide for one known task or alert, such as clearing a full disk or renewing a certificate. A playbook is the wider plan for a type of event, such as a ransomware attack or a major outage: who takes part, which decisions are made and which runbooks are used along the way. A runbook says how to do one thing. A playbook says how to handle the whole situation.

## Key points

- A runbook covers one task or one alert, step by step. A playbook covers a whole event and can call several runbooks.
- The words are not used the same way everywhere. Google SRE calls its alert-response pages playbooks.
- Security teams use "incident response playbook" for a plan by phase: preparation, detection, containment, recovery and review.
- A runbook that nobody has tested since the last system change is often wrong. Date it, test it and tie it to its alert.

## What is the difference between a runbook and a playbook?

A runbook is a set of steps for one job. PagerDuty defines it as "a detailed "how-to" guide for completing a commonly repeated task or procedure within a company's IT operations process, such as provisioning, software updates and deployments, configuration changes, and opening ports" ([PagerDuty](https://www.pagerduty.com/resources/learn/what-is-a-runbook/)). Restarting a stuck queue, rotating a certificate and clearing a full disk are runbook jobs.

A playbook is the plan for a type of event. In PagerDuty's words, "a playbook covers the overall response to larger issues and events, and it can include multiple runbooks and team members as part of the complete workflow" ([PagerDuty](https://www.pagerduty.com/resources/learn/what-is-a-runbook/)). A playbook for a suspected data breach names who leads, who must be told, which decisions need sign-off and which runbooks the responders use.

|  | Runbook | Playbook |
| --- | --- | --- |
| Scope | One task or one alert | One type of event, start to finish |
| Written for | The engineer at the keyboard | Everyone involved, including managers, legal and communications |
| Form | Numbered steps, commands, checks | Roles, phases, decision points, links to runbooks |
| Example | Clear disk space on a database replica | Respond to a ransomware attack |
| Can be automated? | Often, in part or in full | Rarely as a whole; its runbooks may be |


## How do SRE and SOC teams use each term?

Site reliability engineering (SRE) teams keep services running. Google's SRE teams use "playbook" for what many other teams call a runbook. The SRE workbook says "playbooks contain high-level instructions on how to respond to automated alerts" ([Google SRE workbook](https://sre.google/workbook/on-call/)). The first SRE book reports that "recording the best practices ahead of time in a "playbook" produces roughly a 3x improvement in MTTR as compared to the strategy of "winging it."" ([Google SRE book](https://sre.google/sre-book/introduction/)). MTTR means mean time to repair: the average time from failure to fix.

A security operations center (SOC) is the team that watches for and responds to attacks. SOC teams use "playbook" in the wider sense. NIST's incident response guide says "playbooks provide actionable steps or tasks for people to perform during various scenarios or situations" ([NIST SP 800-61r3](https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-61r3.pdf)), and points readers to CISA's playbooks as examples. CISA's federal document "presents two playbooks: one for incident response and one for vulnerability response" ([CISA](https://www.cisa.gov/resources-tools/resources/federal-government-cybersecurity-incident-and-vulnerability-response-playbooks)).

So when two teams argue about the words, check what the document does. If it holds steps for one alert, it works as a runbook, whatever its title says. If it holds roles, phases and decisions, it works as a playbook.


## Runbook example: a full disk on a database replica

The runbook below is an illustration written for this page, not a real company's document. It answers one alert. It says when to stop and hand over to a playbook, and it ends by telling the engineer to record what was done.

Illustrative runbook for one alert:

```markdown
# Runbook: DiskSpaceLow on a Postgres replica
Owner: database team    Last tested: 2026-10-02    Linked alert: db-replica-disk-85pct

1. Confirm the alert: open the disk dashboard for the host named in the page.
   If usage is under 85%, resolve the alert and note "flapped" in the ticket.
2. Find the cause: run  du -sh /var/lib/postgresql/* /var/log/postgresql
3. If the log folder is over 20 GB: compress logs older than 2 days
   (gzip /var/log/postgresql/*.log.2*) and check usage again.
4. If the WAL folder is growing: check replication lag on the dashboard.
   Lag over 10 minutes means the replica is stuck. Stop here and run the
   playbook "Database replica failure" (it pages the database lead).
5. If usage is still over 90% after step 3: grow the volume by 20% with the
   approved Terraform change (module db_replica, variable disk_gb).
6. Record what you did, the commands, and the result in the incident ticket.
```

Three things make it work: a named owner and a "last tested" date, a link to the exact alert, and an exit to the wider playbook when the problem is bigger than the runbook.


## What goes in an incident response playbook?

An incident response playbook is organized by phase rather than by command. CISA's incident response playbook follows "preparation, detection and analysis, containment, eradication and recovery, and post-incident activities" ([CISA playbooks (PDF)](https://www.cisa.gov/sites/default/files/2024-08/Federal_Government_Cybersecurity_Incident_and_Vulnerability_Response_Playbooks_508C.pdf)). For each phase it sets out who acts, what they check and what they record. It also tells responders to "thoroughly document every step taken during this and subsequent phases. Create a timeline of all relevant findings" ([CISA playbooks (PDF)](https://www.cisa.gov/sites/default/files/2024-08/Federal_Government_Cybersecurity_Incident_and_Vulnerability_Response_Playbooks_508C.pdf)).

A SOC playbook for one threat, such as a phishing report, follows the same pattern on a smaller scale: how to triage the report, which logs to pull, when to block a sender or reset a password, and who signs off on closing the case.

| Phase | What the playbook sets out | Runbooks it might call |
| --- | --- | --- |
| Preparation | Contacts, tools, access, severity levels | Grant emergency access; test backups |
| Detection and analysis | How to confirm an incident and judge its scope | Pull sign-in logs; check a host for indicators |
| Containment | Who may isolate systems and when | Isolate an endpoint; block an address range |
| Eradication and recovery | Removing the cause and restoring service | Rebuild a server from a known image; rotate keys |
| Post-incident | Review, lessons learned, follow-up work | Open the postmortem; file action items |


## What should a runbook template include?

PagerDuty cites the seven sections that author Tom Limoncelli "recommends ... for every runbook" of a service, from a service overview to disaster recovery plans ([PagerDuty](https://www.pagerduty.com/resources/learn/what-is-a-runbook/)). For a single-alert runbook, a shorter template does the job:

- Title that names the alert or task, and the alert ID it answers.
- Owner, last-tested date and the systems it applies to.
- Checks that confirm the problem is real before anyone changes anything.
- Numbered steps, each with the exact command or console path and the expected result.
- Exit points: when to stop, whom to page and which playbook takes over.
- Rollback steps for any change the runbook makes.
- A last step that says where to record what was done.


## How do you keep runbooks and playbooks current?

PagerDuty is plain about it: "a runbook isn't something you set and forget. Test and update it regularly so it keeps working as your systems and applications change" ([PagerDuty](https://www.pagerduty.com/resources/learn/what-is-a-runbook/)). It also says "detailed incident reports and postmortems show where a runbook would help most." The best trigger for an update is an incident: if a responder had to improvise, the runbook was missing or wrong.

Keep runbooks in version control or a wiki with history, so each change has an author and a date. Link each one from the alert that pages for it. Over time, move steps into scripts where it is safe: PagerDuty sorts runbooks into three types "based on how much operator involvement they require" ([PagerDuty](https://www.pagerduty.com/resources/learn/what-is-a-runbook/)), from manual to semi-automated to fully automated. For the timeline that shows which runbook was used when, see the [incident timeline template](https://origindatapartners.com/guides/incident-timeline-template).


## Common mistakes with runbooks and playbooks

Most problems show up the first time someone uses the document under pressure.

- Writing a playbook as a wall of commands, so managers cannot follow it and engineers cannot find their step.
- Writing a runbook with no exit, so the engineer keeps trying steps when the event needs a wider response.
- No owner and no tested date, so nobody knows whether the steps still match the systems.
- Pasting passwords, API keys or customer names into the steps.
- Steps that change production with no rollback.
- Keeping the runbook somewhere that goes down with the system it is meant to fix.


## Are runbooks and playbooks also a data asset?

They can be, together with the incidents they were used in. AI data buyers have asked Origin for the records of companies that run and defend their own infrastructure: incidents joined to the change that fixed them, configuration history, ordered logs of what happened, and the runbooks, SOC tickets, on-call records and postmortems from the same period. A runbook on its own is a procedure. A runbook next to the ticket, the chat thread and the fix it led to shows how real people applied it. The records must be about the company's own infrastructure (or a client's, only where that client has agreed), and the company must own them. See the [hub page for infrastructure and incident records](https://origindatapartners.com/guides/license-infrastructure-incident-data-ai), [incident postmortem examples](https://origindatapartners.com/guides/incident-postmortem-examples) and [security incident data](https://origindatapartners.com/guides/license-security-incident-data-ai).

Origin is an independent data partner. Your company decides what is shared and signs any license itself, and a buyer may pay Origin when a deal closes.

- Secrets: runbooks often hold hostnames, keys, tokens and passwords. Scan for them and rotate any you find ([remove secrets from git history](https://origindatapartners.com/guides/remove-secrets-from-git-history)).
- Customer data: remove customer names, tenant IDs and anything copied from a customer environment.
- Personal data: on-call names, phone numbers and emails need masking or consistent pseudonyms.
- Client environments: a managed service or SOC provider's runbooks for a client stay out unless that client has agreed.

Does your company hold records like these? [See what your data is worth](https://origindatapartners.com/check-fit).


## Questions and answers

### What is the difference between a runbook and a playbook?

A runbook gives the steps for one task or one alert, such as clearing a full disk. A playbook is the plan for a whole type of event, such as a ransomware attack: who leads, which decisions are made, who is told and which runbooks the responders use. A playbook often links to several runbooks.

### Is a SOC playbook the same as a runbook?

Not usually. A SOC playbook covers one kind of threat from start to finish, such as a phishing report or a malware alert, with triage, containment and sign-off steps. A runbook is narrower: the exact steps for one task inside that response, such as resetting a user's sessions or isolating a laptop.

### Why does Google SRE call runbooks playbooks?

It is a naming choice. Google's SRE books use "playbook" for the pages that tell an on-call engineer how to respond to an automated alert, which many other teams call runbooks. The SRE workbook describes them as high-level instructions on how to respond to automated alerts.

### What should an incident response playbook include?

An incident response playbook should set out roles and contacts, severity levels, and what happens in each phase: preparation, detection and analysis, containment, eradication and recovery, and post-incident review. CISA's federal playbook follows those phases and tells responders to document every step and build a timeline.

### Can a runbook be automated?

Yes, often in part. PagerDuty sorts runbooks into three types by how much an operator does: manual, semi-automated and fully automated. Simple checks and fixes are good candidates for scripts. Steps that need judgment, such as deciding to fail over a database, usually stay with a person.

### How often should runbooks be reviewed?

Review a runbook after every incident that used it, and after any change to the system it covers. Many teams also set a fixed review date, such as every quarter. A last-tested date on the runbook itself shows readers whether the steps can still be trusted.

## Sources

- [PagerDuty: What is a Runbook?](https://www.pagerduty.com/resources/learn/what-is-a-runbook/)
- [Google, Site Reliability Engineering, chapter 1: Introduction](https://sre.google/sre-book/introduction/)
- [Google, The Site Reliability Workbook, chapter 8: On-Call](https://sre.google/workbook/on-call/)
- [NIST SP 800-61 Rev. 3: Incident Response Recommendations and Considerations for Cybersecurity Risk Management](https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-61r3.pdf)
- [CISA: Federal Government Cybersecurity Incident and Vulnerability Response Playbooks](https://www.cisa.gov/resources-tools/resources/federal-government-cybersecurity-incident-and-vulnerability-response-playbooks)
- [CISA: Cybersecurity Incident and Vulnerability Response Playbooks (PDF)](https://www.cisa.gov/sites/default/files/2024-08/Federal_Government_Cybersecurity_Incident_and_Vulnerability_Response_Playbooks_508C.pdf)

## Related reading

- [Incident Postmortem Examples: Outage, Security and Template](https://origindatapartners.com/guides/incident-postmortem-examples)
- [Incident Timeline Template: Alert to Fix, On-Call Handoff](https://origindatapartners.com/guides/incident-timeline-template)
- [Can a Company License Its Incident and Config Data to AI?](https://origindatapartners.com/guides/license-infrastructure-incident-data-ai)
- [Can a security company license alerts and incident data to AI?](https://origindatapartners.com/guides/license-security-incident-data-ai)
- [Cybersecurity Incident Data: A Guide for MSSPs and IR Firms](https://origindatapartners.com/guides/cybersecurity-incident-soc-case-data)
- [AI SRE Agent Traces: Incident Runs as Training Data](https://origindatapartners.com/guides/ai-sre-incident-agent-traces)
