Summary

  • Record the provider-acknowledged status, service identifier, evidence date and review owner for every locally important telecom dependency.
  • Exercise the named escalation handoff and verify restored voice, data or application service at the user edge before closing the incident.

Start with the service, not the building

A single premises can depend on several services with different owners and failure modes: fixed access, mobile connectivity, inbound call routing, managed wide-area links, cloud authentication and power for customer equipment. “Town Hall” is therefore too coarse an entry.

The bounded unit is a service dependency. Each record should name the operational function it supports, the provider-visible service identifier, the accountable local owner and the evidence supporting any claimed restoration treatment. Sensitive identifiers belong in an access-controlled register, not in a public map or planning paper.

The register must also distinguish three states that are often blurred together:

  1. locally nominated as important;
  2. acknowledged by the relevant provider under a defined arrangement; and
  3. invoked during an incident with a traceable response.

Only the second and third states provide operational evidence. The first is a planning assertion.

Make eligibility an expiring fact

Eligibility and contact routes change. Contracts are renewed, circuits are migrated, accounts are consolidated and buildings change use. A register that was correct last year can be actively misleading during the next outage.

Every entry therefore needs an evidence date, evidence owner and review date. Evidence might be a provider reference, an agreed critical-customer record, a service-management acknowledgement or another documented arrangement applicable to that service. The register should record what was accepted, not what the local team hoped the label meant.

Expiry is a useful control. If evidence is not reconfirmed, the entry falls back to “locally nominated—provider status unverified.” That does not erase the dependency. It prevents an assumption from masquerading as assurance.

Build an escalation chain that can survive the incident

A telephone number is not an escalation chain. The operating record should say who may call, what identifiers must be supplied, which incident threshold applies, how an acknowledgement is captured and where the issue moves if the first route fails.

Cabinet Office guidance places provider liaison and exercises inside local resilient-communications planning. EC-RRG provider guidance also emphasises fault procedures, restoration information and coordination where providers interconnect. The practical local implication is narrow: escalation needs named ownership and a handoff record. It does not mean the local team directs a provider’s network operations.

For multi-provider dependencies, the register should expose the seam. An access circuit may be sold by one organisation and delivered over another network. Call routing may be managed separately. Recording only the commercial account manager leaves the technical handoff invisible.