top of page
40051abd5a76713af8f015988fc6780e-blue-phone-icon-with-a-wave-on-it.webp
Logo-Zalo-Arc.webp
Facebook_Messenger_logo_2020.svg.webp
whatsapp

Co-managed SOC: How to divide roles between the CISO, Internal IT, and MSSP

6 days ago
14 min read

A co-managed SOC is a collaborative cybersecurity operational model where an organization’s internal team and a Managed Security Service Provider (MSSP) jointly operate security operations. Rather than simply splitting alert volume, both parties must clearly establish who monitors, who investigates, who authorizes remediation actions, and who holds ultimate accountability for security risk.

The core distinction between a co-managed SOC and a fully outsourced model lies in the organization's continued operational involvement. The CISO maintains risk governance; the internal security team retains institutional knowledge of business operations, sensitive assets, and systems; IT operations executes infrastructure-level actions; and the MSSP delivers specialized monitoring, analytics, and coverage tailored to agreed-upon requirements.

Consequently, organizations evaluating security service providers should not simply ask which SIEM platform a SOC uses or its operating hours. Far more critical questions include: How is operational scope divided? Who retains data ownership? What level of containment authority does the MSSP hold? And which joint KPIs accurately reflect day-to-day collaborative effectiveness?

What is a Co-Managed SOC?

A co-managed SOC is a shared-responsibility security operations model where the enterprise and an MSSP jointly manage operations based on a predefined scope, explicit authorization levels, and structured coordination protocols. Each party assumes responsibilities aligned with their specific capabilities, access privileges, and environment familiarity.

what-is-a-co-managed-soc
A co-managed SOC is a shared-responsibility operational model where the enterprise and the MSSP share operational duties under a clearly defined scope, authority matrix, and collaboration framework

The MSSP typically monitors security alerts, performs initial triage, investigates early indicators, and escalates verified threats. The internal team oversees log source onboarding, validates business context, coordinates with system owners, and authorizes potentially disruptive remediation actions. The CISO governs overall security direction, defines risk appetite, and enforces operational governance.

This framework differs fundamentally from merely purchasing a SIEM or XDR solution. While modern security tooling centralizes telemetry and triggers alerts, it cannot define accountability, decision-making authority, or escalation workflows on its own. The true differentiator between deploying a SIEM platform and building mature SOC operational capability lies in people, processes, and the ability to translate raw data into decisive action.

Furthermore, a co-managed model does not mean delegating the CISO’s role to an external provider. The NIST Cybersecurity Framework (CSF) 2.0 identifies Governance as a foundational function that establishes, communicates, and monitors cybersecurity risk management strategy, expectations, and policy. Ultimate accountability must always reside within the enterprise's corporate governance structure.

Which organizations are best suited for a Co-managed SOC?

A co-managed SOC is ideal for organizations that have developed foundational in-house security capabilities but lack the resources, headcount, or desire to operate end-to-end monitoring, deep forensic analysis, and 24/7 incident response internally. This need frequently arises when internal IT teams possess deep architectural knowledge of the environment but lack 24/7 coverage, specialized threat hunting expertise, or after-hours coverage.

This model also serves enterprises that require strict administrative control over their tooling, proprietary data, and incident containment decisions. Rather than handing total operational control over to an external provider, the organization can retain management of sensitive use cases, privileged accounts, and mission-critical systems in-house while offloading standardized security operations to an MSSP.

Conversely, a co-managed model will struggle if the enterprise fails to assign designated decision-makers, does not maintain an updated asset inventory, or lacks the bandwidth to handle technical escalations. An MSSP can identify suspicious anomalies, but response efforts stall if no internal stakeholder is available to verify asset criticality, confirm data ownership, or assess business impact.

Before selecting an operational model, organizations should perform a gap analysis comparing existing internal capabilities against their target operating model. Evaluating an in-house SOC versus full outsourcing provides the necessary baseline; a co-managed SOC represents an intentional hybrid approach, not a one-size-fits-all default.

How does a Co-managed SOC differ from In-house and fully outsourced SOCs?

The key distinction among these three models lies in how operational authority and decision-making responsibilities are partitioned. An in-house SOC builds, owns, and manages its own personnel, processes, and technology stack; a fully outsourced SOC delegates the vast majority of day-to-day security operations to a third party; and a co-managed SOC maintains an integrated, collaborative operating framework between both teams.

Model

Enterprise role

Provider role

Key control points

In-house SOC

Owns and operates all people, processes, and technologies

Ad-hoc technology support or specialized consulting as needed

Staffing retention, 24/7 coverage, and ongoing capital/operational costs

Fully outsourced SOC

Manages contract compliance, vendor risk, and high-level approvals

Delivers end-to-end monitoring, analysis, escalation, and scoped response

Vendor lock-in, direct data accessibility, and clear exit strategies

Co-managed SOC

Co-operates the SOC; maintains organizational context and critical decision authority

Augments 24/7 coverage, deep analytical capability, and standardized operational workflows

RACI clarity, shift handovers, containment authority, and shared performance KPIs

The division of labor should not be evaluated solely by the percentage of tasks delegated to the MSSP. An organization may outsource the bulk of repetitive monitoring while still running a true co-managed SecOps model - provided the internal security team actively participates in use case engineering, deep-dive investigations, threat containment, and continuous posture improvement.

Enterprises must also distinguish between SOC as a Service and MSSP. One model emphasizes platform delivery, while the other focuses on managed service operations - though hybrid offerings exist. The commercial label never replaces the need to rigorously examine the granular Statement of Work (SOW).

How can security operations scope be divided?

Security operations can be divided by operational tier, technology layer, security use case, or operational shift. Organizations often blend these models, but each operational workflow must have an unambiguous primary owner and explicit handoff criteria.

Tier-Based Division: The most traditional approach assigns Tier 1 alert triage to the MSSP, Tier 2 investigation to the internal security team, and Tier 3 advanced analysis/threat hunting to infrastructure specialists or the internal CSIRT. While conceptually straightforward, this structure can introduce operational friction if escalation thresholds rely solely on arbitrary "Tier" labels rather than required forensic artifacts.

Technology-Based Division: Suitable for enterprises with mature, specialized internal teams. The MSSP monitors core detection layers (SIEM, EDR, or XDR); internal IT manages physical network infrastructure, endpoints, cloud tenants, and IAM; and specialized application teams manage business-critical services. The primary risk is that multi-stage attacks traverse multiple technology layers, necessitating an appointed cross-functional incident coordinator.

Use-Case-Based Division: Aligns operations directly with specific risk profiles. The MSSP manages high-volume alerts such as phishing, commodity malware, and standard endpoint detections, while internal teams handle insider threats, financial fraud, and proprietary application vulnerabilities. The organization's detection engineering catalog must be built around actual business assets, threat models, and available telemetry, rather than generic off-the-shelf templates.

The MITRE ATT&CK framework documents real-world adversary tactics, techniques, and procedures (TTPs). Organizations can leverage MITRE ATT&CK to map detection use cases and pinpoint defensive coverage gaps, but the framework itself does not dictate which use cases belong to an internal team versus an external provider.

Time-Based Division: Designed for organizations where internal security staff cover local business hours while the MSSP provides overnight, weekend, and holiday coverage (Follow-the-Sun / 24/7 coverage). This structure demands rigid shift-handover protocols, active case tracking, and emergency contact procedures. Without disciplined handovers, alerts generated near shift boundaries risk falling through operational cracks.

How should a RACI matrix be built across the CISO, Internal IT, and MSSP?

A practical RACI matrix must be mapped to granular operational workflows rather than high-level statements like "MSSP monitors and client responds." The matrix must clearly assign who is Responsible (R), Accountable (A), Consulted (C), and Informed (I) for each action.

Bảng dưới đây là mẫu tham khảo:

Activity

CISO

Internal SOC/Security

IT operations

MSSP

System owner

Key deliverable/Output

Log source onboarding

I

A

R

C/R

C

Log source inventory, scope definitions, and health status

Use case development

C/A

R

C

R

C

Detection logic, required data schemas, and acceptance criteria

Alert triage

I

C

I

R/A

I

Initial analysis documentation and triage evidence

Impact verification

I

R

C

C

A/R

Impacted asset details, user identity context, and business impact

Incident investigation

I

A/R

C

R

C

Forensic timeline, working hypotheses, and validated evidence

Endpoint isolation

I

C

A/R

R (if pre-authorized)

I

Approver identity, execution timestamp, and containment results

Account disablement

I

C

A/R

R (if pre-authorized)

C

Target account, justification, and rollback procedures

CSIRT Activation

A

R

C

C

I

Formal activation decision and declared incident scope

Service restoration

I

C

R

C

A/R

Service restoration criteria and functional validation results

Post-incident review

A

R

R

R

C

Lessons learned report and continuous improvement plan

KPI Reporting

A

R

C

R

I

Metric dashboards, trend analysis, and action items

Note: RACI assignments must not be applied rigidly without contextual operational definitions. For instance, an MSSP can only be designated as "Responsible" for endpoint isolation if authorized by contract, technical RBAC, and verified runbooks. Lacking prior authorization, the MSSP must recommend containment and await explicit client sign-off.

The most vital column to define is Accountability. The CISO does not execute technical tasks daily but retains ultimate enterprise risk accountability. IT operations executes many infrastructure changes because they hold native administrative rights and understand operational dependencies. System owners retain authority over whether critical production services can be interrupted or restored.

Who owns the tools and data in Co-managed security operations?

In a co-managed SecOps environment, tooling licenses, data ownership, and operational responsibilities must be decoupled. An MSSP may manage day-to-day operations on a security platform, but this does not grant them ownership of the enterprise's logs, forensic cases, or evidence.

Customer-Hosted Tenant: If security platforms run within the enterprise's cloud tenant or on-premises environment, the organization maintains primary ownership and direct access to raw data. The MSSP is provisioned role-based access control (RBAC) to triage alerts, manage incidents, and optimize detection rules. This model minimizes vendor lock-in but requires internal resources to govern licenses, infrastructure integrations, and IAM policies.

Provider-Hosted Platform: If security telemetry is ingested directly into an MSSP-hosted multi-tenant platform, the contract must explicitly safeguard raw log access, case export capabilities, data export formats, retention windows, and post-termination offboarding terms. An analytics dashboard never substitutes for direct access to raw forensic data required for internal investigations or external compliance audits.

Documentation covering roles and permissions within Microsoft Sentinel demonstrates how granular RBAC distinguishes between viewing data, managing incidents, authoring content, and executing playbooks. This underscores why operational access must follow the principle of least privilege rather than granting broad global admin rights across the entire SOC team.

Beyond baseline access controls, both parties must formally align on data residency, retention policies, audit logging for analyst actions, data usage rights for provider-side model training or platform improvements, and chain-of-custody preservation. The contract's exit strategy must define strict deadlines for data offboarding, supported data export formats, transition assistance SLAs, and the immediate deprovisioning of all provider access accounts.

How should escalation workflows be designed to prevent missed incidents?

A reliable escalation workflow must synthesize threat severity, detection fidelity, business impact, and pre-delegated response authority. A high-severity technical alert may not constitute a high-impact business incident if the targeted asset is non-critical; conversely, a low-fidelity behavioral anomaly warranting rapid escalation may demand immediate priority if associated with a domain controller or privileged identity.

Each escalation severity level must define designated recipients, communication channels, acknowledgment SLAs, minimum required technical artifacts, and clear operational next steps. Standard email notifications may suffice for low-priority events, but severe after-hours incidents necessitate out-of-band communication methods (such as direct phone escalation or dedicated paging tools). Escalation directories must include secondary and tertiary on-call contacts to account for unresponsive primary personnel.

A minimum viable escalation package must contain: impacted asset identifiers and network location, target user accounts and identity context, event timestamps (normalized to UTC and local time), telemetry sources involved, detailed analyst observations and working hypotheses, threat confidence score and potential business impact, initial containment actions already executed.

If an MSSP merely forwards a raw alert screenshot without enriched context, internal teams are forced to repeat initial triage from scratch, defeating the purpose of the partnership.

NIST Special Publication (SP) 800-61 Revision 3 embeds incident handling firmly within enterprise risk management, emphasizing the critical feedback loop between Detect, Respond, Recover, and continuous improvement. This standard underscores that escalation cannot simply be a raw alert handoff; it must directly empower downstream response orchestration and operational recovery.

Enterprises can clarify these operational boundaries by mapping responsibilities across their SOC, CSIRT and CERT, defining precisely where SOC investigative triage concludes and CSIRT incident management takes lead.

What level of containment authority should an MSSP have?

An MSSP should only be granted direct containment authority within explicitly defined parameters, governed by strict trigger conditions, pre-approved asset categories, and tested rollback procedures. Broad, ambiguous contract clauses such as "authorized to mitigate threats" introduce severe operational and availability risks.

  • Notify-Only Mode: The MSSP analyzes telemetry, enriches evidence, and recommends immediate response actions; internal enterprise personnel must review and approve all remediation steps before an endpoint is isolated, an account disabled, or network traffic blocked. While this model guarantees internal operational control, it prolongs Mean Time to Respond (MTTR) if internal contacts fail to respond swiftly.

  • Pre-Authorized Response Mode: The MSSP is pre-authorized to execute targeted containment actions under strict, pre-documented scenarios—such as isolating non-critical staging endpoints or blacklisting verified external malicious IPs at the perimeter. Dedicated runbooks must detail minimum confidence thresholds, excluded business-critical assets, operational escalation time limits, real-time stakeholder notification pathways, and unambiguous rollback runbooks.

Security automation must operate under the same permission boundaries. Microsoft Sentinel, for instance, utilizes automation rules to trigger playbooks based on granular conditions, but the platform itself requires explicit permissions to execute logic apps against target resources. The technical capability to automate an action does not mean that every action should run unattended.

For mission-critical production systems where disruption incurs severe business consequences, system owners or IT operations must retain final approval authority. Organizations should conduct regular tabletop exercises to test not only containment speed, but also service restoration procedures.

What cadence should be established for Co-managed SOC governance?

An effective co-managed SOC governance model operates across three distinct operational cadences: daily/weekly tactical coordination, monthly operational service reviews, and quarterly strategic reviews. Each cadence addresses a specific tier of operational decision-making.

Daily/Weekly Tactical Touchpoints: Focus on active high-priority cases, critical unresolved alerts, broken log pipelines, pending remediation tickets, and shift-handover bottlenecks. The primary objective is removing operational friction before incident backlogs grow or detection coverage gaps persist.

Monthly Service Reviews (MSR): Evaluate contractual SLAs, operational KPIs, false-positive ratios, log ingestion health, escalation quality, and detection engineering/tuning backlogs. Reporting must analyze root causes rather than merely reciting raw metrics. For example, if Mean Time to Detect (MTTD) spikes because a critical firewall log feed dropped silently, the corrective action required is entirely different from an analyst-driven triage delay.

Quarterly Strategic Reviews (QBR): Address high-level cyber risk posture, enterprise architecture changes, detection use case roadmaps, tabletop exercise outcomes, budget forecasts, and internal security maturity. These sessions serve as the venue to review provider dependency, track bidirectional knowledge transfer progress, and adjust the commercial SOW as enterprise needs evolve.

The CISO maintains ultimate decision authority over risk priorities and enterprise risk acceptance. The MSSP provides objective threat data, telemetry analytics, and technical recommendations; the internal security team contributes organizational context; and executive leadership authorizes the necessary capital and operational resources for major strategic shifts.

What shared KPIs should be tracked between the enterprise and MSSP?

Shared KPIs must evaluate telemetry coverage, detection engineering quality, operational response velocity, and cross-team collaboration. Relying exclusively on high-level metrics like MTTD or MTTR obscures the underlying operational bottlenecks slowing down incident resolution.

  • Coverage Metrics: Track onboarded log sources, telemetry pipeline stability, critical asset monitoring coverage, and active detection use cases. A SOC that processes alerts quickly still leaves the enterprise exposed if mission-critical crown jewels fail to forward logs.

  • Quality Metrics: Track false-positive rates per detection use case, triage completeness, evidence quality in escalation tickets, and missed detections identified during post-incident reviews or purple team engagements. Avoid targeting zero false positives at all costs; overly aggressive alert tuning often suppresses high-value malicious signals.

  • Response Velocity Metrics: Dissect incident lifecycles into granular stages: Mean Time to Acknowledge (MTTA), Mean Time to Triage, Mean Time to Escalate, Decision-Making Latency, Mean Time to Contain, and Mean Time to Recover. While the MSSP drives triage and escalation velocity, containment duration frequently hinges on internal approval workflows and IT operations execution.

  • Collaborative Operational Metrics: Track the volume of cases awaiting internal client review versus cases awaiting MSSP investigation, escalations rejected due to missing technical context, actions violating the RACI matrix, and overdue detection improvement tasks. These metrics are the most critical in a co-managed model, as operational outcomes depend entirely on synchronized cross-team execution.

  • Outcome-Driven Metrics: Ultimately, metrics must reflect real-world security outcomes: repeat incident frequency, use cases developed or tuned following an incident, resolved detection gaps, and performance during incident simulation drills. Never rely on a single isolated metric to assign blame when operational success requires the alignment of technology, personnel, and institutional authority.

What data should enterprises prepare prior to an assessment?

Before initiating an assessment with an MSSP, an enterprise must compile a comprehensive inventory of assets, network architecture, available log sources, deployed security tools, active detection use cases, historical incident reports, and internal team capabilities. Lacking these baseline inputs, providers are forced to build operational scopes and commercial proposals around unverified assumptions.

  • Log Source Inventory: Detail all core systems, designated system owners, log output formats, daily ingestion volumes (EPS/GB per day), retention requirements, and existing collector integrations. The organization should also document its deployed SIEM, EDR, XDR, SOAR, and IT Service Management (ITSM) ticketing tools, alongside current licensing tiers.

  • Operational Parameters: Define desired monitoring coverage windows (e.g., 24/7/365 vs. after-hours), operational time zones, escalation paths, technical response permissions, prioritized threat use cases, and strict lists of sensitive assets excluded from automated containment. Incident response logs from previous security events help external providers understand common alert types, historical bottlenecks, and internal forensic evidence requirements.

  • Commercial Framework: Clearly establish which components constitute baseline operational services versus consumption-based metrics billed by data ingestion volume, monitored asset/user counts, ticket volumes, or dedicated professional service hours. When comparing vendor proposals, two offerings can only be evaluated objectively if evaluated against an identical baseline of tools, telemetry sources, coverage hours, and containment scope.

What boundaries must be defined in a Co-managed SOC contract?

A robust co-managed SOC contract must clearly delineate service scope, SLA structures, tooling ownership, data governance, administrative access rights, containment limits, and termination/offboarding workflows. Vague contractual statements such as "provider will monitor and resolve security incidents" fail to provide enforceable operational baselines or clear legal boundaries.

  • Statement of Work (SOW): Must detail specific log sources, supported use cases, operational coverage windows, communication languages, primary escalation channels, and formal reporting deliverables. SLAs must delineate acknowledgment, triage, and escalation times from downstream actions that depend on client-side internal approvals.

  • Data Governance Clauses: Must establish explicit enterprise ownership over all ingested security data, defined geographic storage locations, retention minimums, data export rights, extraction formats, and post-termination data sanitization protocols.

  • Access and Credential Management: Must dictate designated analyst accounts, mandatory Multi-Factor Authentication (MFA), administrative audit logging, regular privilege reviews, and immediate deprovisioning workflows upon analyst reassignment or termination.

  • Exit Strategy and Offboarding: Contracts must mandate the orderly transition of active cases, custom detection rules, use case documentation, API integration lists, and historical operational data necessary to transition security operations in-house or to a new partner. If underlying tools are hosted within the provider's ecosystem, the contract must define which developed assets (such as custom rules or playbooks) are intellectual property belonging to the enterprise.

  • Legal and Risk Terms: Liability caps, security breach notification SLAs, subcontractor transparency, third-party audit rights, and financial service-level credits must be reviewed by legal, procurement, information security, and enterprise risk management teams. This guidance does not constitute formal legal counsel.

Should organizations start with a pilot or a full deployment?

Organizations should begin with a structured pilot deployment whenever log data quality, detection use cases, the RACI matrix, or response authorization models have not yet been validated in a live operational setting. A direct full-scale deployment is only appropriate when an enterprise already possesses an established internal operational foundation and is simply migrating to a new partner.

A pilot must be scoped to be representative without becoming unwieldy. The optimal approach selects a subset of critical business assets, high-fidelity log sources, and stable detection use cases with robust underlying data. Furthermore, the pilot phase must intentionally test both routine operational alerts and high-severity escalation procedures.

Acceptance criteria must extend beyond the raw number of alerts closed. The evaluation should focus on data completeness, triage precision, handoff efficiency, the speed of identifying internal decision-makers, and the safe execution of containment runbooks.

Following the pilot, both teams must refine the RACI matrix, tune detection logic, update documentation, calibrate shared KPIs, and formalize the long-term commercial scope. A co-managed security operations model should only expand organization-wide after cross-team coordination has been validated operationally, not merely because security tools have established an API connection.

How can IPSIP Vietnam assess and build a Co-managed SOC model?

Our engagement methodology begins with an exhaustive assessment phase designed to evaluate critical business assets, existing log sources, and your current security tool investments across SIEM, EDR, and XDR platforms. Based on your internal team's capabilities and identified operational gaps, IPSIP Vietnam collaborates directly with your stakeholders to define scope boundaries, build the RACI matrix, design escalation workflows, establish CISO approval thresholds, and configure core shared KPIs.

ipsip-viet-nam
IPSIP Vietnam delivers comprehensive assessments and builds tailored co-managed SOC architectures alongside end-to-end enterprise cybersecurity solutions

Following the initial assessment, both teams initiate a structured pilot across representative enterprise assets and high-priority use cases. This phase validates raw telemetry quality, alert classification accuracy, cross-team decision latency, and authorized threat containment capabilities prior to enterprise-wide rollout.

Depending on each enterprise's unique security architecture, IPSIP Vietnam delivers customized cybersecurity solutions - ranging from SOC 24/7 monitoring and hybrid MSSP models to advanced detection and response (XDR, NDR) and specialized Incident Response. Co-managed SOC pricing is determined transparently following the assessment, mapped precisely to ingested log scope, tool integrations, coverage tiers, and data ownership models, ensuring optimal security spend aligned with real-world enterprise risk.



References

NIST Cybersecurity Framework 2.0 - National Institute of Standards and Technology

MITRE ATT&CK - MITRE


follow ipsip vietnam.png
bottom of page