Before the dial tone dies: a seven-question resilience audit for UK voice estates in 2026

The UK's analogue phone network is entering its final stretch. Openreach's current plan is to retire the public switched telephone network by the end of January 2027, and in 2026 that date should be treated as an operational deadline rather than a supplier milestone. The risk is not the desk phone. It is the line behind the lift, the alarm, the payment terminal and the contact-centre queue.

Many organisations have a telecoms inventory. Few have a dependency map. The difference matters because a PSTN sunset audit is less about circuits and more about business continuity, compliance and the workflows that assume a dial tone will always be there.

1. What is actually hanging off the line?

Start with a physical and contractual survey, not a spreadsheet exported from the telecoms provider. Build a line-level register that records circuit identifiers, site location, connected device, business owner, contract end date and criticality. Include the lines nobody claims to own: door-entry systems, fax machines in finance, legacy modems, building-management links and standby lines that were installed for a reason now forgotten. Landlords, facilities teams and security contractors often hold part of the answer. The audit should also distinguish between lines that can move to an all-IP service and devices that need replacement or vendor certification before they can run over broadband.

2. Which dependencies are safety-critical, regulated or reputationally exposed?

Not every analogue line carries the same risk. Lift emergency phones, fire-alarm signalling, intruder alarms, telecare and nurse-call systems, payment terminals and lone-worker devices need a higher tier of scrutiny. Some must remain on monitored, resilient connections rather than a standard broadband router. Engage health and safety, fire safety, clinical governance and finance teams early. Ask suppliers for written confirmation that their equipment is supported on the proposed replacement, and test the answer rather than accepting a sales brochure. A voice migration that leaves a lift phone unable to call for help is not a telecoms project issue; it is a safety and continuity failure.

3. What is the right migration sequence?

A big-bang cutover is rarely defensible in a dispersed UK estate. Sequence by risk, site, supplier readiness and business cycle. Standard office telephony may move to SIP trunking or a cloud calling platform first, with parallel running for a short period. Contact centres need more careful handling because call flows, interactive voice response, recording, workforce management and CRM integration all interact. Alarm and lift lines may need a separate track with specialist suppliers. The sequence should show dependencies, not just dates: which sites need new connectivity, which numbers must be ported, which devices need replacement and where staff training must happen before the old service is removed.

4. Will resilience survive the transition?

Voice over IP is only as reliable as the power and connectivity beneath it. The audit should test failure modes, not just successful calls. What happens during a broadband outage, a power cut, a SIP provider incident or a cyber event? Are there diverse links, cellular failover and battery backup for critical devices? Can staff still reach the 999 service, and does the emergency call carry accurate location information? Contact centres should test rerouting to secondary sites or home workers. Facilities teams should test alarm signalling end to end. Document recovery objectives and incident playbooks, then run a live exercise. Resilience that exists only in a supplier assurance document is not resilience.

5. Are supplier exit terms and service levels fit for the next estate?

Legacy contracts often reflect a world of copper lines and long notice periods. Before migration, review termination charges, migration assistance, number-porting responsibilities, service credits and data extraction rights. Who owns call recordings and configuration data? What happens if a port fails? Does the contract auto-renew while a migration is underway? Ask for a clear exit plan and named escalation contacts. The cheapest line rental can become the most expensive dependency if leaving requires months of notice or a specialist migration fee. Procurement should treat exit terms as part of the business case, not an afterthought once the new platform is chosen.

6. Who owns the change, and how will adoption be proven?

PSTN retirement cuts across IT, facilities, security, operations, finance and, in some sectors, clinical or care teams. A single owner at executive level is needed, with a working group that can make cross-budget decisions. Success should be measured by service outcomes rather than lines removed: alarm test pass rates, call answer rates, abandoned-call trends, payment success, staff ticket volumes and first-contact resolution. Adoption matters because a technically complete migration can still fail if staff do not trust the new call handling or if customers cannot reach the right queue. Report progress in business language, not circuit counts.

7. What happens the day after the switch?

The audit does not end when the last legacy line is ceased. Decommission old equipment, update the asset register and remove forgotten dependencies from supplier contracts. Put periodic testing into business-as-usual routines, especially for safety-critical devices. Keep an incident playbook for voice outages and review it after every test or live event. Revisit the dependency map when buildings are refurbished, tenants change or new payment and security systems are installed. The organisations that manage this well will treat the 2026 audit as the start of continuous assurance, not a one-off compliance exercise. For ReChannel readers, the lesson is clear: the dial tone is disappearing, but the dependencies behind it need to be re-channelled into a managed, tested and accountable service.