Outsourcing can give a growing business access to skills, coverage and operational discipline it could not reasonably maintain alone. The difficulty is rarely that every provider is doing nothing. More often, each provider is delivering its part of the agreement while an important responsibility between them remains unclear.
One supplier manages devices, another hosts an application and a third provides cloud services. When something fails, who is actually expected to investigate, authorise a response or restore the business service? I would rather resolve those questions during an ordinary service review than discover the answer during an incident.
Replace broad assurances with specific responsibilities
“The MSP handles security” is not an operational description. Does that include monitoring, and during which hours? Who investigates an alert? Who has authority to isolate a device? What happens if the nominated customer contact cannot be reached?
Cloud services also involve a division of responsibilities, which varies according to the product and contract. The supplier may operate the platform while the business remains responsible for account configuration, access approval or retention decisions. A platform availability commitment does not, by itself, establish how an essential business service will recover.
I would use a straightforward responsibility matrix for significant activities. Record the accountable business owner, the party undertaking the work, the escalation route and the evidence expected. If authority is delegated in advance—for example, permission to disable a suspected compromised account under agreed conditions—that needs to be explicit too.
The phrase 'the supplier handles security' is particularly unhelpful when several parties deliver different parts of a service. Monitoring, investigation, technical containment, business approval and customer communication may sit with different teams. I would start with actions rather than generic responsibility labels. Who receives the alert, who verifies it, who can isolate a device and who can authorise a disruption to operations? Then check whether those people actually have the access, availability and contractual authority required. A responsibility matrix should support a real decision at an inconvenient hour, not only satisfy a governance review.
Test the hand-offs
Consider an illustrative out-of-hours alert affecting a finance account. The monitoring company sends a warning. The managed service provider has the technical access to disable the account, but the contract requires customer approval. Nobody can reach the authorised contact. Each party has followed its procedure, yet the business may be left exposed while everyone waits.
The remedy isn't necessarily another tool. It may be a deputy contact, clearer escalation timing and an agreed decision about authority to act. That agreement should consider both sides of the problem: the damage caused by delay and the disruption a mistaken intervention might create.
The NCSC's supply-chain security guidance encourages organisations to understand supplier risk, establish appropriate control and review whether arrangements work [S8]. A proportionate approach concentrates attention on suppliers that could materially affect important business services.
The most revealing question in an exercise is often 'what happens next?' A monitoring provider may deliver a notification correctly while no one sees it until the following morning. A technical team may have the ability to act but be unable to contact the authorised customer representative. Neither failure is fixed by collecting more security alerts. Exercises should include the awkward exceptions: an absent contact, a disputed alert severity or a containment action that risks interrupting an important transaction. Those discussions help decide which powers can safely be delegated and where an alternative contact or escalation route is essential.
Ask for evidence that matches the service
A supplier's certification can provide worthwhile assurance, but only about its stated scope. It does not necessarily show how a particular customer environment has been configured or whether a specific recovery arrangement will work as expected.
The most helpful evidence is often quite focused: an example escalation record, the actual coverage of a monitoring service, the outcome of a recent restoration or a review of privileged accounts. Each should answer a real management question. Requesting a large pack of policies without a clear purpose can consume time without reducing uncertainty.
Agree the evidence requirements early and revisit them when the service changes. There may be reasonable limits on what a provider can disclose. Where that happens, the business needs to judge whether alternative evidence is adequate and record any remaining gap.
Evidence requests should be narrower and more useful than asking every supplier for a large policy pack. If out-of-hours response is the concern, ask which systems are covered, who receives the alert and for a representative example of how notification became action. A certification may be helpful within its stated scope but cannot automatically establish the configuration and recovery capability of a particular customer environment. Where a supplier cannot share operational records because they contain other customers' information, leadership should consider alternative assurance and explicitly record whatever uncertainty remains.
Include failure and exit
Supplier resilience deserves attention too. What if the provider itself becomes unavailable? Can the business obtain essential records, contact the right people and recover administrative control? Are there subcontractors or shared platforms on which several services quietly depend?
Exit arrangements are best discussed before they become urgent. Decide how business data will be retrieved, which configuration records and credentials must transfer, who is responsible for those steps and how access will be removed afterwards. The detail belongs in the relevant technical and commercial agreements, not just in a reassuring paragraph about portability.
Independent advice can help where the parties have different expectations. Its role is to clarify the business requirement, supporting evidence and practical options, while leaving contractual interpretation and specialist technical work with the appropriate professionals.
Provider exit is a resilience question, not only a procurement task. A service may work very well while leaving the business dependent on a single team for credentials, configurations or usable data. An orderly exit plan should identify what can be retrieved, who can grant access to a replacement provider and how services will continue during handover. Confirmation that data can be exported is not the same as proof that it can be restored or used. A limited rehearsal, conducted while the relationship is healthy, can expose missing information before a transition becomes urgent.
Sources and further reading
[S8] NCSC. Supply chain security guidance. Read original guidance
[S1] UK Government. Cyber Governance Code of Practice. Read original guidance
The source collection records these references as checked on 28 October 2026. Source attribution does not imply endorsement.
Scenarios are illustrative, not client case studies. This article provides general business guidance, not a security audit, certification or substitute for sector-specific technical, legal or contractual advice.
