Enterata All articles
Digital Operations

Green Is Not Evidence: Testing UK Supplier Incident Communications

Enterata
Green Is Not Evidence: Testing UK Supplier Incident Communications

Photo: BLM Nevada, Public domain, via Wikimedia Commons

There is a familiar ritual during a supplier outage. The status page turns amber, then green. A short update says the team is investigating, then monitoring, then resolved. The service returns. The incident is closed. For many UK organisations, that sequence becomes the official record of what happened. It should not.

A status page is a communication channel. It is not a monitoring system, an audit log or a statement of customer impact. A green indicator proves that a supplier chose to display green. It does not prove that the service was available to the customer, that data was preserved, that failover worked, or that the recovery time matched the contract. Teams that treat vendor updates as evidence are outsourcing their operational memory to the party with the strongest interest in a tidy narrative.

The green page problem

Status pages are designed for reassurance at scale. They summarise complex incidents into a small number of states. That summarisation is useful during an outage, but it erases the details that matter afterwards: when the fault began, which customers were affected, which regions or tenants were involved, whether data was lost, and what workaround was offered.

The problem is not that suppliers lie. The problem is that a status page is not built to answer the questions a UK operations team will face in a post-incident review, an insurer conversation or a board paper. It is built to reduce inbound support traffic. Those are different objectives.

Treat incident communication as a control to test

A more defensible approach is to treat supplier incident communication as a control. Controls are tested. They have expected behaviour, evidence and exceptions. For each critical supplier, the organisation should know how it will be notified, how quickly, through which channel, and what information the supplier is contractually obliged to provide.

Test the channel before an incident. Ask the supplier to send a test notification through the same route used for major incidents. Confirm who receives it, whether it reaches an on-call rota or a shared mailbox, and whether the message is actionable. A notification that arrives in a mailbox nobody monitors is not a control.

Questions worth asking a supplier

A useful supplier conversation moves beyond assurances and asks for specifics. When did you detect the incident? How did you detect it? Which services and tenants were affected? Was any data lost, corrupted or exposed? What failover was invoked, and did it meet the stated recovery objective? What workaround was available, and when was it communicated? What was the customer-facing impact, as distinct from the internal technical impact? What evidence will be included in the post-incident report?

These questions should be asked before renewal, not only during an outage. The answers reveal whether the supplier can produce evidence at all. A supplier that cannot distinguish detection time from notification time, or customer impact from internal severity, is unlikely to provide a defensible account when something goes wrong.

Keep an internal incident evidence log

The organisation should maintain its own record of supplier incidents, even when the supplier provides a report. The log does not need to be elaborate. It should capture the time the organisation first noticed the issue, the time the supplier acknowledged it, the time the supplier first communicated customer impact, the status page updates observed, the internal decisions taken, the workarounds used, the recovery time observed by the organisation, and any discrepancies with the supplier's account.

That log serves three purposes. It creates an operational memory that survives staff changes. It gives the organisation a way to challenge a supplier narrative with contemporaneous evidence. And it provides the raw material for contract discussions, insurance claims and board assurance.

Contract clauses to revisit before the next outage

Many supplier contracts mention notification but do not define it precisely. Before the next renewal, UK teams should check whether the contract states the notification timeline, the channel, the content required and the consequences of late or incomplete notification. It should also clarify whether the status page is the sole notification method. It should not be.

Other clauses worth revisiting include the right to receive a post-incident report within a defined period, the right to audit or review evidence, the definition of service availability, the treatment of scheduled maintenance, the customer's right to communicate with affected downstream users, and the relationship between service credits and actual operational loss. Service credits may compensate for a billing period. They rarely compensate for the cost of a failed month-end close.

Make the post-incident review evidence-led

A post-incident review should compare the supplier's account with the organisation's own log. Where the two differ, the review should record the difference rather than resolve it by default in favour of the supplier. That is not adversarial. It is the minimum standard for learning from an incident that crossed an organisational boundary.

For Enterata, the practical conclusion is simple. Green is a colour, not a conclusion. UK teams should test the communication path, keep their own record, and make suppliers show their working. The next outage will be easier to manage if the evidence was designed before it happened.

All Articles

Related Articles

Cited but Unclaimed: Auditing the Dashboards That Shape UK Decisions Without an Owner

Cited but Unclaimed: Auditing the Dashboards That Shape UK Decisions Without an Owner

Who Holds the Pager? A Pre-Holiday Cover Audit for UK Critical Platforms

Who Holds the Pager? A Pre-Holiday Cover Audit for UK Critical Platforms

Nobody Ever Resigns: Putting Machine Identities Through the UK Leaver Process

Nobody Ever Resigns: Putting Machine Identities Through the UK Leaver Process