- A NOC watches for outages and slowdowns. A SOC watches for attacks. Many providers sell both, but they are different services.
- The provider usually works in tiers: automation first, then a first-line engineer with runbooks, then a deeper second line.
- The SLA should state response and resolution times by priority, and what the provider may change without asking.
- Your tickets, alerts and change records describe your network. Make sure the contract lets you export them.
What is NOC as a service?
A network operations center (NOC) is the team that keeps networks and systems up. PagerDuty describes a NOC as "the centralized location (in-office or virtual) for an organization's network team", a team that "monitors networks, servers, application infrastructure, cloud usage, and more for events that could degrade or disrupt service for customers and users" (PagerDuty).
NOC as a service means another company runs that team for you. It is a form of managed service. A joint advisory by CISA and six partner agencies defines managed service providers (MSPs) as "entities that deliver, operate, or manage ICT services and functions for their customers via a contractual arrangement, such as a service level agreement" (CISA AA22-131A). ICT means information and communications technology. Canada's cyber agency notes that "an MSP may have its own network operation center (NOC) to ensure that the customer's IT operations run smoothly" (Canadian Centre for Cyber Security).
The model suits companies whose systems must stay up overnight and at weekends but who cannot staff three shifts. For what an in-house NOC does day to day, see the network operations center guide.
What does a NOC as a service provider do?
PagerDuty lists the usual NOC duties, starting with "providing 24/7 coverage, typically across two or three shifts" and ending with "notifying the team responsible for a given service when an outage occurs and the NOC cannot solve it" (PagerDuty). An outsourced NOC does the same work for several clients from one floor.
| Task | What the provider does | What stays with you |
|---|---|---|
| Monitoring | Watches devices, links, servers and cloud services for faults and slowdowns | Deciding what is monitored and at what thresholds |
| Ticketing | Opens, updates and closes a ticket for each event | Access to the ticket history and the right to export it |
| First-line fixes | Restarts services, fails over links, clears known faults with runbooks | Approving the runbooks and any change outside them |
| Escalation | Calls your engineers or vendors when it cannot fix the fault | Keeping the contact list and on-call rota current |
| Reporting | Uptime, ticket counts and SLA results each month | Reviewing the reports and acting on repeat faults |
Most providers staff in tiers. PagerDuty describes level 0 as automation that "should be fully automated and act as the first line of defense before a human ever sees an issue". Level 1 engineers "work through attached, predetermined runbooks to resolve routine, documented issues", and "if an L1 can't resolve a problem with a runbook, the L2 handles further triage and troubleshooting using deeper expertise" (PagerDuty). The runbooks are the heart of the service; see runbook vs playbook.
What is the difference between a NOC and a SOC?
A NOC keeps services available. A security operations center (SOC) looks for attacks. Canada's cyber agency draws the line by provider type: a managed security service provider (MSSP) "establishes one or more security operations centers (SOCs) which is responsible for monitoring and protecting the security of their customer's infrastructure", while an MSP's NOC exists so that IT operations run smoothly and the MSP "may not provide security related monitoring as part of that service" (Canadian Centre for Cyber Security).
| NOC as a service | SOC as a service | |
|---|---|---|
| Main question | Is it up and fast? | Is someone attacking it? |
| Typical alerts | Link down, high latency, disk full, failed backup | Suspicious sign-in, malware on a laptop, data leaving the network |
| Main tools | Network and server monitoring, ticketing, runbooks | SIEM, endpoint detection, threat intelligence |
| Success measure | Uptime and time to restore service | Time to detect and contain an attack |
The two overlap in practice. A failing firewall is a NOC ticket; a firewall rule changed by an intruder is a SOC case. If you buy both, ask how the two teams pass work to each other. The security side is covered in SOC as a service.
What should a NOC as a service SLA and contract include?
The service level agreement (SLA) is the part of the contract that sets measurable targets. The UK National Cyber Security Centre (NCSC) gives a starting point for managed services: "For general service requests or minor issues, a response time of 1 business day is standard. For urgent issues, a response time of under 1 hour is standard." It defines response time as "the time between logging an issue and your MSP starting to investigate it" (NCSC). It also says "a matrix of responsibilities detailing what the MSP will do and what is expected from you as the customer is good practice", and lists health reports that include "monitoring and uptime statistics (servers, network, cloud services)" (NCSC).
CISA adds records terms. Its checklist for MSP customers asks for "detailed guidelines for log and records maintenance" and "specific performance-related service level agreements, including a clear delineation of operational IT services and security services" (CISA Insights). That second point matters when one provider runs both your NOC and your security tools.
- Priority levels with examples, and the response and restore target for each.
- Hours of cover, and whether weekends and holidays count toward the targets.
- Which changes the provider may make on its own, and which need your approval.
- How tickets, alerts, change records and reports are stored, for how long, and how you export them.
- What happens to that history, and to the provider's access, when the contract ends.
- Service credits or other remedies when targets are missed.
When does NOC as a service make sense?
It tends to fit when outages outside office hours are costly but the network is not large enough to justify a full night shift. It fits less well when the network is unusual, changes daily or needs engineers who know the business closely.
| Situation | Usually points to |
|---|---|
| Many branch sites, standard equipment, small IT team | NOC as a service |
| Customer-facing systems that must be watched at night | NOC as a service or a follow-the-sun in-house team |
| A carrier or hosting network that is the product itself | In-house NOC, perhaps with overflow support |
| Frequent custom changes and home-built systems | In-house team, or a provider that co-manages with your engineers |
Ask how the provider prices the service (per device, per site or a flat fee), what counts as out of scope, and how many other clients each engineer covers at night.
Example: one NOC ticket from alert to close
The ticket below is an illustration written for this page, not a real company's record. A fictional chain of 30 clinics uses a NOC provider. One clinic loses its main internet link at night.
| Time (UTC) | Event | Who |
|---|---|---|
| 02:14 | Monitoring flags the primary link at clinic 12 as down (gateway 192.0.2.1 not responding). Ticket NOC-4471 opens as priority 2. | Automation (level 0) |
| 02:15 | Traffic has already failed over to the backup cellular link. The ticket notes reduced bandwidth. | Automation |
| 02:21 | Level 1 engineer follows runbook "Branch primary link down": checks the router, confirms the carrier circuit is down. | NOC level 1 |
| 02:34 | Engineer opens a fault with the carrier and adds the carrier case number to the ticket. | NOC level 1 |
| 06:50 | Carrier reports a cut cable, repair due by 11:00. Ticket updated; clinic manager emailed before opening. | NOC level 1 |
| 10:42 | Primary link restored. Traffic moves back. Engineer confirms normal latency for 30 minutes. | NOC level 2 |
| 11:15 | Ticket closed with cause "carrier cable cut", time on backup 8 h 28 min, no SLA breach. | NOC level 2 |
Every row has a time, an actor and a reference. That is what makes a ticket history useful later, for a monthly report, a dispute with the carrier or an incident timeline.
Common mistakes when buying NOC as a service
Most problems trace back to what the contract left unsaid.
- Assuming the NOC also watches for attacks. Unless the contract says so, it does not.
- Agreeing response times without restore targets, so a ticket can be "responded to" for days.
- Letting the provider write runbooks nobody on your side has read or approved.
- No rule on which changes the provider may make on its own.
- Tickets kept only in the provider's system, with no export right and no copy on your side.
Who owns the NOC tickets, and are they a data asset?
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 and routing history, ordered logs of what happened, and the runbooks, tickets, on-call records and postmortems from the same period. A NOC ticket history is close to that, but whose it is depends on whose network it describes.
A NOC provider's tickets mostly describe client networks. The request covers a company's own infrastructure, and a client's only where that client has agreed. Contracts usually keep the client in control of its own data: CISA tells MSP customers to keep "controls and logs separate from the vendor's records" (CISA Insights), and where personal data is involved, GDPR requires the provider to process it "only on documented instructions from the controller" and, at the client's choice, to delete or return it when the service ends (GDPR Article 28). General information, not legal advice. In short, the client, as the owner of the network and its records, is the one who can decide.
So the natural holders are the client company, for the tickets about its own network, and the provider, for the records of its own internal systems and tooling. Operators that run their own large networks should also see the ISP network operations guide and the hub for infrastructure and incident records. 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: tickets about a client's network stay out unless that client has agreed in writing.
- Secrets: device passwords, SNMP strings and keys pasted into tickets come out, and are rotated.
- Personal data: engineer and site-contact names, phone numbers and emails are masked or replaced with consistent pseudonyms.
- Addresses: public IP addresses and hostnames are replaced consistently (anonymize logs and captures).
Does your company hold records like these? See what your data is worth.
Common questions
What is NOC as a service?
NOC as a service is an outsourced network operations center. A provider monitors your network, servers and cloud services around the clock, opens tickets for faults and slowdowns, fixes routine problems with agreed runbooks and escalates the rest to your engineers or vendors. You pay a subscription instead of staffing your own night and weekend shifts.
Does NOC as a service include security monitoring?
Not unless the contract says so. A NOC watches availability and performance: links, devices, servers and backups. Watching for attacks is the job of a security operations center. Some providers sell both, but they are separate services with separate tools and staff, so check the scope in writing.
What response time should a NOC as a service SLA promise?
UK NCSC guidance treats a response within one business day as standard for minor issues and under one hour for urgent ones. A good NOC SLA goes further: it sets a target per priority level for both response and restore time, says which hours count and names the remedy when targets are missed.
Who owns the tickets an outsourced NOC creates?
That depends on the contract. The tickets describe the client's network, so most contracts treat them as client records held by the provider. Check that the contract gives you a right to export tickets, alerts and change records during the term and a copy at the end, before the provider deletes them.
Can a NOC provider make changes to my network without asking?
Only as far as the contract and the approved runbooks allow. Most agreements let the provider take routine, documented steps, such as restarting a service or failing over a link, and require approval for anything else. Write the boundary down, and make sure every change the provider makes is logged in the ticket.
How is NOC as a service usually priced?
Providers commonly price by the number of devices or sites monitored, by service tier or as a flat monthly fee. Ask what counts as out of scope, whether after-hours work costs extra, how many clients each engineer covers at night, and what the service credits are when the provider misses its targets.
Sources and further reading
- PagerDuty: Network Operations Center Best Practices
- CISA and partners: Protecting Against Cyber Threats to Managed Service Providers and their Customers (AA22-131A)
- Canadian Centre for Cyber Security: Choosing the best cyber security solution for your organization (ITSM.10.023)
- UK National Cyber Security Centre: Choosing a managed service provider (MSP)
- CISA Insights: Risk Considerations for Managed Service Provider Customers (PDF)
- GDPR Article 28: Processor
