Practitioners who have done this almost never report losing a client over it. Clients update a contact and carry on. What people actually lose is a few weeks of stray calls — someone dials the old number in a bad moment and gets silence, or gets a stranger.

So the whole job is three things, and none of them are technical.

Keep something answering the old number for a while. Not forever. Long enough that nobody hits a dead line during the changeover.

Tell everyone once, in writing, with a date. One message, the new number, the day it takes over. Not a mention at the end of a session that half of them will forget by Thursday.

Find every place the old number is written down. This is the part that eats the time, and the part people underestimate. The number is on more surfaces than you remember putting it on.

If you also want to keep the old number rather than replace it, that is a different mechanism — a transfer, not a switch — and it changes the sequence quite a bit. There is a section on that below.

What actually goes wrong

Not clients leaving. Four other things.

The gap. The old line stops working before the new one is properly out. Every person who calls in that window gets nothing, and you never find out who they were. This is the only failure in the list that costs you an actual person.

The number is written somewhere you forgot. One practitioner described cancelling a service that hadn't worked out and being left with the old number already printed on business cards she had handed out. Paper does not update. Neither does a directory listing you set up three years ago and never logged back into.

You keep giving out the old one. The most common post-switch failure is not clients ignoring the new number. It is the practitioner still handing out the old one from habit — on an intake form, in an email signature, in a voicemail greeting recorded before the change.

Your callback shows an unfamiliar number. A therapist who forwarded an after-hours line to a cell described calling a client back in a crisis and having them see a different number and get confused. The change is invisible to you and very visible to them.

Should I move the old number over, or start a new one?

Two genuinely different projects.

Starting fresh is simple and immediate. You pick up a new number, run both for a while, and retire the old one on your own schedule. Nothing can fail. The cost is that everyone has to learn a new number.

Transferring the old number — porting it to a new provider — means nobody has to learn anything, because the number does not change. The cost is that it is a process with other companies in it, it takes time, and it can fail.

Transferring is worth it when the number is on printed material or has been in circulation for years. Starting fresh is worth it when the number is young, when it is attached to a service you're glad to be rid of, or when the number itself is part of the problem — a line drowning in spam is not worth carrying with you.

The mechanics of a transfer — what it needs from you, how long it takes, and how it fails — are their own subject, and worth reading before you commit to one.

What the switch actually looks like

In order, because the order is what prevents the gap.

1. Get the new line working before you announce anything. Call it. Text it. Leave yourself a voicemail and check that it arrives. Have someone else call it from a different network. Do this before a single client knows it exists.

2. Record the new greeting. Whatever you want the line to say when you can't answer, have it in place on day one. A new number with a default carrier greeting sounds like a wrong number.

3. Set the old line to point at the new one. A forward, or a greeting that states the new number slowly and twice. The goal is that nobody who dials the old number is left with nothing.

4. Send the message. Once, in writing, to everyone — the new number, the date it takes over, and a direct ask to save the contact.

5. Update the written surfaces. The list below.

6. Wait, then retire the old line. Not the week after. Long enough that the stragglers have surfaced.

Where the old number is actually written down

Work this as a list, not from memory. In rough order of how often people miss them:

• Your email signature

• Your directory listing, and any second directory you forgot you were on

• Your website — including the footer and the contact page, which are often edited separately

• Your intake forms and consent documents

• Your invoices and receipt templates

• Your voicemail greeting on the other line

• Your out-of-office auto-reply

• Any business listing that shows up when someone searches your name

• Printed material — cards, a brochure, a poster in a shared waiting room

• The referral sources who have it saved: family doctors' offices, other therapists, an EAP, a school

• Your own outgoing message templates, if you have any saved

The referral sources are the ones with the longest tail. A GP's office that has your old number written on a sticky note will keep using it for years.

What it costs to leave it alone

If the current number works, staying is free and correct. Switching is only worth doing when the number itself is the problem — it rings your personal phone, it is on a service you have lost confidence in, it is buried in spam, or it belongs to someone else in a way that worries you.

If you are unhappy with a provider, check whether you can take the number with you before you decide anything. That single fact changes the whole plan.

The short version

The switch is a communications project with a small telecom errand inside it. Overlap the two lines, send one clear written message with a date, and work the list of places the old number is written down. Do that and the clients are fine — they always are. It is the sticky note in a doctor's office that takes a year.

About Rivet

Rivet is a Canadian practice line built for therapists — a separate number for calls, texts and voicemail, with video sessions and clinical tools in the same place. Your data stays in Canada, and transcription runs on Rivet's own hardware rather than a third-party AI service.

One plan, $65 CAD a month, everything included. Fourteen-day trial, no card.

See what's included · Start the trial