- A SOC detects and responds to attacks that got past your other defenses. The service model rents that team and its tools.
- Outsourcing buys coverage and skills. It costs business context, which you must supply.
- Agree in writing what the provider may do on your systems without asking, and how fast it must tell you.
- Keep direct access to the logs and case history. Seven agencies from five countries advise keeping the most important logs for at least six months.
What is SOC as a service?
A security operations center (SOC) is the team that watches for and responds to attacks. The UK National Cyber Security Centre (NCSC) puts it this way: "The role of a SOC is to limit the damage to an organisation by detecting and responding to cyber attacks that successfully bypass your preventative security controls" (NCSC).
SOC as a service (SOCaaS) means a provider runs that team for you, from its own facility and usually on its own tools. The US federal government buys it this way too. CISA's catalog lists a shared service run by the Department of Justice that "delivers 24x7x365 threat monitoring, detection and incident response, threat intelligence, and cybersecurity investigations to customers" (CISA).
The engine of most SOCs is a SIEM (security information and event management) tool. The NCSC describes it simply: "It takes logs and data sources as input, performs some correlation and rule checking, and then outputs alerts for triaging" (NCSC blog). Analysts then decide which alerts are real.
What does SOC as a service include?
Packages differ, so read the service description line by line. The federal service on CISA's list includes items such as "Logging-as-a-Service", "Managed Security Information and Event Management (SIEM)", endpoint detection and response, forensics and a "Client portal for metrics and dashboards" (CISA). A commercial offer usually covers some mix of the following.
| Component | What it means | Question to ask |
|---|---|---|
| Log collection | Your firewall, identity, cloud and endpoint logs are sent to the provider | Which sources are included, and what does adding one cost? |
| 24/7 monitoring and triage | Analysts review alerts and sort real ones from noise | How many alerts are closed without a person looking? |
| Investigation | Analysts work out what happened, on which systems | What evidence will you show me for each finding? |
| Response | Containment steps, such as isolating a laptop or disabling an account | Which steps may you take without asking me first? |
| Threat hunting | Searching for attacks that raised no alert | How often, and what do I see from it? |
| Reporting | Case summaries, monthly metrics, a portal | Can I export every case and its notes? |
Is SOC as a service better than an in-house SOC?
Neither is better in general. The NCSC notes that "most SOCs will outsource some of their operations, even if it is just the signatures that your tools will use for detection". It lists the main benefit of in-house work as "better contextual awareness of the business, which will aid any investigations", and says it "can also be effective to outsource a function if it is seldom used, or if the SOC simply doesn't have the resources to implement the function internally" (NCSC).
Canada's cyber agency names the main limit of outsourcing: "Unlike an in-house IT team, an MSSP may not always be aware of the new services and changes that your organization implements." It also reminds buyers that "you are putting your security in outsiders' hands while you ultimately own the risk" (Canadian Centre for Cyber Security).
| Model | Fits when | Watch for |
|---|---|---|
| Fully outsourced | Small or mid-sized security team, need for night and weekend cover | Thin business context; slow hand-offs to your IT team |
| Co-managed (hybrid) | You have analysts in office hours and want cover outside them | Unclear ownership of a case that spans both teams |
| In-house | High-risk sector, unusual systems, enough staff for shifts | Hiring and keeping analysts; tool costs |
What should you ask a SOC as a service provider?
Canada's cyber agency publishes 17 questions for evaluating a managed security provider. Several go to the heart of a SOC contract: "Ask for an example SLA and review it to see if it responds to your requirements in terms of speed of detection, alerting, and resolution", "Do you have access to any of the audit data?" and "Is the client given access to the MSSP's systems and interfaces (dashboarding/APIs/accounts)?" (Canadian Centre for Cyber Security).
A joint advisory from CISA and its partners adds two that buyers often miss. It says "it can be months before incidents are detected" and recommends that organizations "store their most important logs for at least six months". It also says "contracts should detail how and when MSPs notify the customer of an incident affecting the customer's environment" (CISA AA22-131A).
- Which log sources are in scope, and how long each is kept.
- Notification times by severity, and who on your side is called at 3 a.m.
- The response actions the provider may take alone, written as a list.
- Where your logs are stored, in which country, and who at the provider can read them.
- How you export logs, cases and analyst notes, during the contract and at the end.
Who keeps the logs and case notes in SOC as a service?
General information, not legal advice. CISA's guidance for managed service customers asks for "direct access to security logging information, network intrusion detection, and anomaly analysis data telemetry from all systems managed by the MSP", and says customer policies should include "maintaining controls and logs separate from the vendor's records" (CISA Insights). Where logs hold personal data, as most do, a provider acting for an EU or UK client is a processor under GDPR: it acts "only on documented instructions from the controller" and, "at the choice of the controller, deletes or returns all the personal data to the controller after the end of the provision of services" (GDPR Article 28).
In practice the raw logs come from your systems, and the case notes are written about your systems in the provider's tools. Ask before signing whether you get both back, in what format, and how long the provider keeps copies after the contract ends. Related reading: audit log examples.
Example: one alert handled by a SOC as a service
The case below is an illustration written for this page, not a real company's record. A fictional logistics company with 300 staff uses a SOC provider whose contract allows it to disable accounts but not to change firewall rules.
| Time (UTC) | Event | Who |
|---|---|---|
| 21:03 | SIEM rule fires: a dispatcher account signs in from 203.0.113.45, a new country, after 14 declined MFA prompts. | SIEM |
| 21:07 | Analyst confirms the MFA prompt was approved on the 15th try. Case SOC-2291 opens as high severity. | Provider, level 1 |
| 21:12 | Analyst disables the account and revokes its sessions, as the contract allows. Client on-call is called. | Provider, level 1 |
| 21:30 | Mailbox audit shows a new forwarding rule to an outside address. Rule removed; finding added to the case. | Provider, level 2 |
| 21:48 | Client blocks the address range at the firewall, a step the provider may only recommend. | Client IT |
| Next day 09:10 | Dispatcher confirms the prompts were not theirs. Password reset, MFA changed to number matching. | Client IT |
| Next day 11:00 | Case closed: credential theft, contained in 9 minutes, no data sent out. Lessons logged. | Provider and client |
The split between provider and client actions is set by the contract. The case only reads as one story because both sides wrote to the same case and timeline; see the incident timeline template.
Common mistakes with SOC as a service
These show up most often in the first real incident after signing.
- Sending the provider only a few log sources, then assuming everything is watched.
- No written list of response actions, so the provider waits for a call while the attack spreads.
- An out-of-date contact list, so the 3 a.m. call reaches someone who left.
- Leaving out business context: new systems, travel, planned changes. Analysts then chase false alarms.
- Logs and case notes kept only on the provider's platform, with no export right.
- Treating the service as a full incident response retainer. Many SOC services contain an attack but do not rebuild systems.
Are SOC records a data asset, and whose are they?
AI data buyers have asked Origin for the records of companies that defend their own infrastructure: incidents joined to the change that fixed them, ordered logs of what happened and what was done, and the SOC tickets, runbooks, on-call records and postmortems from the same period. A closed SOC case with its alert, triage, actions and outcome is that kind of record.
A SOC provider's cases come from client environments, and the request covers a company's own infrastructure, a client's only where that client has agreed. The contract terms above point the same way: logs are the client's to keep separate and to get back, and a processor works on the client's documented instructions. The client, as the owner of the systems and the records, is the one who can decide. A company that buys SOC as a service is therefore the natural holder of its own cases, provided it can export them. A provider can offer the cases from its own corporate environment. For the provider's view, see cybersecurity incident and SOC case data and the security incident data guide. 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.
- Client environments: cases from a client's systems stay out unless that client has agreed in writing.
- Secrets and live weaknesses: credentials, keys and unfixed vulnerabilities come out first.
- Personal data: user names, emails and IP addresses are replaced with consistent pseudonyms (anonymize logs and captures).
- Shared intelligence: indicators received from others under sharing limits stay out.
Does your company hold records like these? See what your data is worth.
Common questions
What is SOC as a service?
SOC as a service is an outsourced security operations center. A provider collects your security logs, monitors them around the clock, triages alerts, investigates the real ones and either tells you or takes agreed containment steps when it finds an attack. You pay a subscription instead of hiring and keeping analysts for every shift.
What is the difference between SOC as a service and an in-house SOC?
An in-house SOC is your own staff, with deep knowledge of your systems and business. SOC as a service rents a provider's analysts and tools, which buys round-the-clock cover and specialist skills but less business context. Many companies mix the two, keeping analysts in office hours and using a provider at night.
What logs should I send to a SOC as a service provider?
Start with the sources attacks pass through: identity and sign-in logs, endpoint detection, firewalls and VPN, email security and cloud control-plane logs such as administrative actions. Then add critical servers and applications. Agree retention for each source; CISA and partner agencies advise keeping the most important logs for at least six months.
Can a SOC as a service provider take action on my systems?
Only as far as the contract allows. Many agreements let the provider take containment steps such as isolating a laptop or disabling an account, and leave firewall changes and system rebuilds to the client. Write the allowed actions down as a list, with who must be told and how fast.
Who owns the logs in a SOC as a service contract?
Usually the client, because the logs come from its systems, but the contract decides. CISA advises customers to keep direct access to logging data and keep their own records separate from the provider's. Under GDPR, a provider processing personal data acts on the client's instructions and deletes or returns it when the service ends.
Is SOC as a service worth it for a small business?
It can be, when attacks outside office hours are a real risk and the company cannot staff its own analysts. It is less useful if few log sources are connected or nobody on your side can act on the provider's calls. Compare the scope, response actions and export rights, not only the monthly fee.
Sources and further reading
- UK National Cyber Security Centre: Building a Security Operations Centre (SOC)
- CISA: Security Operations Center as a Service (SOCaaS)
- UK National Cyber Security Centre: To SOC or not to SOC?
- UK National Cyber Security Centre: Building a SOC, designing an operating model
- Canadian Centre for Cyber Security: Choosing the best cyber security solution for your organization (ITSM.10.023)
- CISA and partners: Protecting Against Cyber Threats to Managed Service Providers and their Customers (AA22-131A)
- CISA Insights: Risk Considerations for Managed Service Provider Customers (PDF)
- GDPR Article 28: Processor
