Multi-Site PSTN Migration: Moving Many Lines Across Several Buildings

Quick answer

A multi-site PSTN migration is not one big version of a single-site job - it is a project, and it needs a plan rather than a series of phone calls. The sequence that works: build one estate-wide inventory of every line and charge, survey connectivity at each address (expect a mixed picture), migrate one pilot site properly, then move the rest in waves grouped by risk rather than by geography. The savings are real and usually larger than businesses expect, but they come from consolidation and decommissioning, not from the headline per-user price. The biggest risk is not technology - it is losing track of lines nobody owns.

I have sat on the finance side of enough of these to know where the money goes and where it hides. On a single site, the worst case is a bad afternoon. Across eight or eighteen sites, the same disorganisation becomes months of parallel running, duplicate charges, and a long tail of lines still being billed at three sites you thought were finished. The good news is that the same scale that creates the risk creates the opportunity: this is usually the first time anyone has looked at the whole estate's telecoms at once, and what that audit finds routinely pays for a meaningful chunk of the project. If you would like us to run the inventory and price the estate, get a no-obligation quote.

Why multi-site is a different job, not a bigger one

Four things change once you have more than one building:

Ownership goes fuzzy. On a single site, someone knows what the lines do. Across an estate, lines get inherited with premises, added by site managers, and paid centrally by someone who has never visited. This is why the audit matters so much.

Availability differs. Full fibre at head office, SoGEA at the depot, and something awkward at the unit on the industrial estate. Your connectivity plan is per address; only the voice platform can be uniform.

Numbers get complicated. Each site typically has its own geographic area code and possibly its own DDI ranges. Porting has to be planned per number range, per losing provider, and the details have to match old account records that may be years out of date.

Contracts are staggered. Different sites signed different terms at different times, and some will carry early termination charges. That is a commercial planning problem, and it usually shapes the running order.

Step 1: one inventory for the whole estate

Build a single spreadsheet before you speak to any supplier. For every line at every site, capture:

That last column is the one that pays for the project. Pull three months of bills, list every standing charge, then have somebody at each site physically walk the building and match sockets to charges. Every estate-wide audit I have been involved in has found lines that nobody could account for: fax lines, lines for card terminals that moved to mobile data years ago, lines at premises that were given up. Because legacy line rental rose in stages through 2026 - see the WLR price rises - those lines are more expensive now than they have ever been, so cancelling them is immediate margin.

Mark anything you cannot identify as "unknown" rather than assuming it is spare. Ceasing a line that turns out to have been feeding a lift is a bad way to learn what it did.

Step 2: connectivity survey, address by address

Voice will ride over the internet connection at each site, so each site needs its own answer. Check what is actually available - full fibre, SoGEA, or in some cases a leased line where the site genuinely warrants it. Our SoGEA vs FTTP comparison covers how to choose, and the short version is: take full fibre wherever it exists.

Two commercial points. Leased lines have much longer lead times, often three to four months and sometimes more, so if any site needs one, that site sets your overall timeline and you needed to start it already. And do not upgrade connectivity at a site you might close - check the property plans before ordering a three-year circuit.

Step 3: pilot one site properly

Resist the temptation to start with the easiest site. Pick one that is representative: it should have a real mix of users, at least one non-phone device, and connectivity typical of the estate. The point of a pilot is to surface problems, and an easy site surfaces none.

Run it end to end - inventory, connectivity, port, device migration, cutover, testing, then cease. Then write down what went wrong, because it will go wrong the same way at the other sites. What you are really building is a repeatable runbook: a checklist somebody can follow at site fourteen without you in the room.

Allow a couple of weeks of parallel running before ceasing anything at the pilot. It costs a little in duplicate charges and saves far more in confidence.

Step 4: waves grouped by risk, not geography

The instinct is to migrate region by region. Group by risk and dependency instead:

Wave 1 is the one people get wrong. Lift and alarm work runs on other people's engineers, with lead times commonly four to twelve weeks and lengthening as the deadline approaches - see lift emergency lines and alarm signalling. Those calls should be made in week one across the whole estate, in parallel with everything else, because nothing you do can compress them.

Step 5: numbers and a proper number plan

This is where a multi-site migration earns its keep, because you can design something coherent instead of inheriting an accident.

• Port every range, per site, and check the account details on the port request match the losing provider's records exactly. Mismatched names and addresses are the commonest cause of rejected ports, and each rejection costs you a slot.
• Keep the geographic numbers. Local numbers at local sites matter commercially, and they port fine.
• Design internal dialling once - a consistent extension scheme across the estate rather than four historical schemes.
• Decide on a single main number with routing, if it suits how customers actually contact you.
• Never cease before the port completes. At estate scale, this needs to be a rule with a named owner, not an intention.

Inter-site calling becomes free on a single platform, which for businesses whose sites call each other constantly is a genuine line-item saving. See keeping your business landline number for the porting mechanics.

Step 6: devices, per site, in writing

Every site gets its own device list, and somebody physically confirms it. Lifts, monitored alarms, card terminals, door entry, fax, telemetry, and anything else that dials out. Our devices that rely on phone lines guide is the checklist.

Two rules that save money and embarrassment. Get one estate-wide quote from your lift and alarm companies rather than site-by-site pricing - the per-unit cost usually improves. And do not let the site that nobody visits be the site nobody surveys; in my experience that is exactly where the forgotten lift line lives.

The contract problem, and where the savings actually are

Staggered contract end dates are the main commercial constraint. You have three options, and the right answer is usually a mix: let short-dated contracts run to term and migrate at renewal; pay early termination charges where the saving plus the avoided legacy price rises exceeds them; or negotiate the exit as part of a consolidated deal with the incoming provider, who has a clear interest in helping.

Do the arithmetic per site rather than as a policy. A site three months from renewal should almost always wait. A site twenty months into a legacy contract at 2026 line-rental prices frequently should not.

Then the savings, which for a multi-site business are usually the strongest part of the business case:

• Decommissioning lines the audit proves are unused - immediate, and often substantial.
• Consolidating many site-level contracts into one, with volume pricing and one bill to check.
• Free inter-site calls on a single voice platform.
• Removing copper line rental entirely, at 2026 prices.
• Right-sizing capacity on concurrency rather than the channel blocks each site was sold - see multi-line phone systems and channels.
• Retiring on-site hardware where it is out of support, though keep it where it is sound - see can you keep your existing phone system.

Our switch-off cost guide covers the one-off spend that sits against these.

One owner, or it drifts

The single best predictor of whether a multi-site migration goes well is whether one named person owns it. Not a committee, and not "the IT manager will pick it up alongside everything else". They need the inventory, the authority to cease lines, and a standing slot with whoever is delivering it.

Give that person three artefacts: the estate inventory, the pilot runbook, and a wave plan with dates. Track two numbers weekly - sites migrated, and lines ceased. The second is the one that gets forgotten, and it is the one that determines whether you actually stop paying.

The bottom line

Multi-site migration is a project-management exercise wearing a telecoms costume. Get one inventory, survey every address, pilot a representative site, then move in waves ordered by risk with the lift and alarm calls made across the whole estate in week one. Keep a named owner and track lines ceased as carefully as sites migrated, because parallel running that nobody ends is how a good project quietly becomes an expensive one.

If you would rather hand the whole thing over, this is exactly the work we do - inventory, availability checks per address, wave plan, porting and cutover: get a no-obligation quote or call 0333 006 9399.

Getting unknown calls on a business line? Our free UK phone number checker shows which network or provider Ofcom allocated any number to, and whether the range is in service — no sign-up.