Openreach has set 31 January 2027 as the retirement date for the legacy Public Switched Telephone Network (PSTN), with BT’s ISDN switch-off following the same timetable. The deadline was extended from 2025, but Openreach’s July 2026 update confirmed there would be no further extension.
For IT directors reviewing business phone systems, that does not automatically mean replacing the entire telephony environment. The main options are to retain an existing PBX with SIP trunking, move call control to a hosted PBX, or adopt UCaaS as part of a wider communications refresh.
This guide compares those approaches, the migration work involved and what to assess when choosing a provider.
What changes with the PSTN switch-off?
The transition replaces traditional exchange-based phone services and ISDN circuits with IP alternatives.
But the impact can extend beyond voice. Broadband services and connected equipment may also depend on legacy lines, so audit the whole connection rather than just the handset.
Check every site for:
- Phone lines and numbers
- Broadband dependencies
- Lift phones
- Alarm systems
- Payment terminals
- Fax machines
- Other connected equipment
Also confirm the underlying network operator at each location. Openreach’s January 2027 timetable does not automatically apply to every network.
Full fibre is not required for every replacement. SOGEA, for example, provides broadband without a traditional voice line over Openreach’s fibre-to-the-cabinet network.
Once the estate is clear, the next decision is what replaces the existing telephony service.
UCaaS, SIP trunking and hosted PBX: three different choices
VoIP describes the underlying technology used to carry voice over IP.
The architecture around it determines who controls calling, which applications are included and how much existing infrastructure remains.
UCaaS: replace the wider communications platform
Unified communications as a service (UCaaS) combines calling, video meetings, messaging and collaboration in a cloud-delivered platform. It can replace an existing PBX alongside separate communications tools.
UCaaS fits a broader communications refresh where the organisation wants to consolidate services. Identify which existing systems would be retired and which would remain, then include both in the cost model.
SIP trunking: retain the PBX, replace its external connection
Session Initiation Protocol (SIP) trunking connects compatible telephony infrastructure to a provider over IP. It replaces the legacy external connection while leaving an existing PBX in control of calls.
Check:
- PBX compatibility
- Remaining support life
- Required gateways
- Network configuration
- Failover
- Troubleshooting responsibilities
SIP trunking can support a phased migration or retain a PBX that still meets business requirements.
Hosted PBX: move telephony without replacing everything else
A hosted PBX moves call-control functions to provider-hosted infrastructure without necessarily replacing the organisation’s wider collaboration tools.
Typical functions include:
- Call routing
- Voicemail
- Transfers
- Extensions
- Queues
- Reception features
Check exactly what is included and which functions require additional licences.
| Architecture | When to consider it | Cost categories to compare | Migration work to examine |
| UCaaS | A wider communications refresh, including distributed teams | User subscriptions, calling plans, contact-centre licences, integrations, devices and training | Identity setup, application changes, call-flow rebuilding and user adoption |
| SIP trunking | Retaining a supported PBX that meets requirements | Trunk capacity, calls, PBX maintenance, gateways, connectivity and support | PBX compatibility, routing, firewall configuration, porting and failover |
| Hosted PBX | Cloud-managed telephony alongside existing collaboration tools | User or call-path subscriptions, calling allowances, devices, recording and integrations | Recreating extensions, queues and reception functions; testing devices and workflows |
Compare the full contract cost, including implementation, overlapping services, administration and eventual exit.
Where Microsoft Teams Direct Routing fits
Microsoft Teams Direct Routing connects Teams Phone to a telephony provider through a certified session border controller (SBC).
It is a connectivity option within Teams Phone rather than a separate architecture alongside UCaaS, SIP trunking and hosted PBX.
If Teams is part of the proposed design, check:
- Teams Phone licensing
- SBC requirements
- Carrier connectivity
- Existing telephony dependencies
- Local network requirements
- Troubleshooting ownership
Compare Direct Routing with Operator Connect, Microsoft Calling Plans and other UCaaS options.
Direct Routing can support specialist routing requirements or retained telephony infrastructure. The important question is who owns each part of the service when something fails.
Plan migration backwards from January 2027
In July 2026, Openreach estimated that around 1.5 million legacy lines remained, including services at roughly 350,000 business premises. That makes provider capacity, engineering availability and number porting part of the migration risk.
Legacy services are also becoming more expensive. Openreach increased Wholesale Line Rental pricing during 2026, with another increase scheduled for 1 October 2026.
For organisations still migrating, the remaining work falls into four stages.
September-October: audit, design and secure delivery dates
Start with the asset register, then reconcile it against telecoms bills and physical checks.
Record every number, line, site, contract and connected device, including direct-dial ranges, lift phones, alarms and payment terminals. Assign an owner to each service and identify any broadband dependencies.
Use that picture of the estate to select the replacement architecture. Before finalising the plan, obtain written dates for connectivity, configuration, equipment delivery and number porting.
October–November: pilot and prepare users
Test the new service alongside the existing system before moving live numbers. Include reception staff, customer-service teams and remote users so the pilot reflects different calling requirements.
Test transfers, queues, voicemail, caller identification, integrations and representative busy periods. Validate emergency calling and location information with the provider.
Use the pilot to identify changes to call flows, devices and user processes before the wider migration. Users should also know how to complete routine tasks and what to do if calling fails.
November-December: move in controlled stages
Where supported, move numbers in agreed batches rather than treating the estate as one cutover.
Define the acceptance criteria, business communications, escalation contacts and fallback routing for each stage. Document how parallel running will work and what recovery options remain once a number has been ported.
Confirm those arrangements with both the existing and replacement providers. A live number cannot always be switched back immediately after porting.
January: resolve exceptions and complete handover
Keep the final weeks for unresolved services, testing and handover rather than routine migration work.
Before cancelling a legacy service, confirm that the replacement is working, connected equipment still operates and any dependent connectivity will remain in place.
Openreach’s business guidance says unmigrated businesses may move to emergency voice-only access, EVAc, or lose service altogether.
Treat that as a last-resort network measure, not a business continuity plan.
Score providers on evidence, not feature counts
Once you have chosen an architecture, give each shortlisted VoIP provider the same requirements and test scenarios. That makes it easier to distinguish demonstrated capability from a long feature list.
Score each criterion from 0 to 5, where 0 means no evidence and 5 means validated fit backed by contractual commitments. Apply the weighting to produce a total score out of 100.
| Criterion | Weight | Evidence to request |
| Migration and number porting | 20 | Named delivery owner, validated number inventory, porting dependencies, cutover plan and escalation process |
| Availability and recovery | 20 | Service-level agreement (SLA), uptime definition, exclusions, restoration targets, backup routes and demonstrated failover |
| Integration and usability | 20 | Working tests with your customer relationship management (CRM), helpdesk and identity systems; usable reception and administration workflows |
| Security and information handling | 15 | Administrative access controls, fraud controls, recording permissions, retention options and data-location information |
| Contract and exit | 15 | Contract length, full-term costs, renewal provisions, licence reductions, termination terms, data export and number-transfer arrangements |
| Support ownership | 10 | Support hours, incident escalation, responsibilities across suppliers and reporting after service failures |
Use the score to structure comparison, not to override critical requirements. Emergency calling and safety-critical compatibility should be treated as pass/fail: an unresolved dependency cannot be offset by stronger scores elsewhere.
For integrations, ask providers to demonstrate the full workflow rather than confirming that an integration exists. An incoming call should identify the customer, open the correct record and log the outcome in your customer service platform. Establish who maintains that connection when either system changes.
Also separate platform availability from site resilience. Digital telephony still depends on local power, connectivity and network equipment. Test backup power and an alternative route for calls as part of the design. A cloud service can remain available while an individual site loses the ability to make or receive calls.
Choose the operating model, not just the replacement line
The PSTN switch-off sets the deadline, but it should not dictate the architecture.
Start with the services and workflows the organisation needs to protect. Then compare UCaaS, SIP trunking and hosted PBX against the existing estate, migration requirements, ongoing responsibilities and full contract cost.
The next step is to test those requirements with potential providers. The Elevate.Tech Summit allows senior IT leaders to discuss current technology projects through pre-arranged one-to-one meetings with solution providers.
If PSTN migration, UCaaS or a wider communications refresh is on your roadmap, register your interest and tell the team what you are reviewing so it can inform your meeting selections.
Frequently asked questions
What happens to my landline number after the switch-off?
You can normally retain it when moving to digital telephony. Ask the receiving provider to validate every number and direct-dial range before scheduling the transfer. Keep the existing service in place until the providers have confirmed the migration and cancellation arrangements.
Do I need SIP trunking if I’m moving to UCaaS?
Not necessarily as a separately purchased service. Check how the platform provides external calling: carrier connectivity may be included or arranged separately. Retaining SIP trunks or using Direct Routing can support a hybrid design, but neither is a universal prerequisite for UCaaS.
What happens to fax machines and alarm systems?
Check each device with its manufacturer, maintainer and communications provider. Equipment may need reconfiguration, an adaptor or replacement. Successful voice calls alone do not establish compatibility: confirm end-to-end operation and power-failure behaviour before removing the existing connection.
How long does migration typically take?
There is no single timescale for every business. Multiple sites and connected equipment add work and testing dependencies. Ask for a schedule covering connectivity, configuration, pilot testing, number porting and acceptance, with responsibility assigned to each stage.
Image credit: https://unsplash.com/photos/black-and-gray-ip-phone-WQqViZ_Zd8M
