How it works
From a blank page to a finished recording
Most people are done in under ten minutes. Here is what each step asks of you, what it does for you, and exactly what you are holding at the end of it.
How it works
Five steps, start to finish
Nothing here asks you to understand audio. You choose what you want it to say and how it should sound; the studio handles the rest.
01Type
Tell us what it is for
On-hold, a welcome greeting, an after-hours message, a menu or voicemail. The studio adjusts to suit.
02Script
Write it, or have it drafted
Write your own words, or let the assistant draft them from your business name or your website, then edit anything you like.
03Voice
Pick an Australian voice
Every voice is listed with how it sounds and the sort of business it suits, so the choice takes a minute rather than an afternoon.
04Music
Set the shape of it
How long the recording runs, and how much room sits between one message and the next. Levels and timing are handled for you.
05Preview
Hear five seconds, then decide
A five-second preview of your own audio, in the voice you picked. Pay only when it sounds right.
The words
Three ways to get a script
This is the part that takes the longest and the part worth spending the time on. Start wherever suits you — every route ends in the same editor, with the words yours to change.
Write it yourself
Type it, or paste the script you already use. Words from an old recording go in as they are, and nothing has to pass through the assistant.
Have it drafted for you
Say who you are and what the recording is for, and the writing assistant drafts it. Shorten it, warm it up or rewrite one line without starting again.
Take it from your website
Give us your web address and we read the page for your hours, services and location, then draft from that. Nothing is read until you press the button.
As you write, the studio counts the words and shows how long the finished recording will run, at a comfortable telephone reading pace. It is the readout that stops a two-minute message quietly becoming a six-minute one.
If your business name, a surname or a suburb is often said wrong, select it and write it the way you would explain it to a new receptionist — Fotias is said FOH-tee-us — and the voice will follow you.
How it sounds
Choosing the voice
12 Australian voices — 6 female and 6 male. Every one is generated by AI, not a recording of a voice actor. Each is described by how it sounds and the sort of business it suits.
In the studio every voice plays the same short demo line, so you are comparing like with like rather than judging one voice on a sentence that happened to suit it. Filter to the female or male voices if that decision is already made, and change your mind as often as you like before you pay.
Recordings are produced voice-only while the music library is closed, and the studio does not ask you to choose a bed you cannot have. The levels, the timing between messages and a clean start and finish are all still set for you. Nothing further is asked of you.
Meet the voicesBefore you pay
Five seconds of your own audio
The preview is your words, in the voice you chose, with the music you chose, mixed the way the finished recording will be. It is ready about a minute after you ask for it.
Five seconds is the whole of what we make before payment. The complete recording is not produced and then withheld — it is not produced at all until your payment goes through. Five seconds is enough for the two things that actually decide it: whether that voice suits your business reading your words, and how the music sits underneath them.
Heard something you would change? Change it and listen again. A line that reads awkwardly, a voice that is not quite the business, music that is too busy behind the words — take a new preview each time, and buy when one of them sounds right.
After you pay
Pay once, then download
One payment for the recording you make. No subscription, no monthly fee and nothing to cancel. Your complete recording is produced within a few minutes of payment and the downloads appear on the same page.
A file for your phone system
Prepared for 3CX, Yeastar, Grandstream, or a general file if you use something else, with the upload steps beside it.
MP3
Plays anywhere. Good for sharing, checking on your phone, or email.
Master WAV
The full-quality original. Use this if someone asks for the highest quality file.
Downloads are unlimited and the script you approved is kept beside the recording, so whoever loads it can read what it says without playing it. You do not need an account to buy. Afterwards you can sign in with an emailed link to keep your recordings together and reach them from another computer.
Straight answers
The four things people want to know
Will it sound robotic?
The voices are generated rather than recorded by a person, and we say so plainly. What usually gives phone audio away is the mix rather than the voice: dead air between sentences, levels that jump, a loop that clicks on the join. That part is handled for you. Judge the rest on your preview, where the voice you chose is reading your own script rather than a line we picked.
What if I do not like it?
Change the script, the voice or the music and take another preview, as often as you like. Nothing is charged until you have heard it and said yes, and you cannot check out without a preview behind you. If a recording you have already bought is not right, email us and we will sort it out with you.
Do I need to know anything technical?
No. You are never asked about file formats, levels or settings. You choose your phone system from a list, or choose Other if you are not sure what sits in the comms cupboard, and the file you download is prepared to suit it. If your provider looks after the system for you, send them the file and they will load it.
How long does this really take?
Most people are finished in under ten minutes. Writing is the slow part, which is why the assistant will draft something for you to edit. A preview is ready about a minute after you ask for it, and the complete recording is produced within a few minutes of payment.
Worth knowing
What to put in a good on-hold message
A caller on hold is a captive audience with a low tolerance for being sold to. The recordings that work answer the question the caller rang about, and then get out of the way.
- Answer the question they rang about
- Opening hours, where you are, how to book, what to bring, how long a quote takes. Most callers ring about one of five things. Say those five things and the recording has earned its place.
- Put the useful part first
- Plenty of callers are answered inside a minute, so the opening twenty seconds do most of the work. Lead with what they need, not with how long you have been in business.
- Keep each message short
- Two or three sentences, then let the music carry a gap. A caller cannot re-read a line they missed, and a long paragraph is harder to follow down a phone line than on a page.
- Give them something to do besides wait
- A booking page, an email address, an after-hours number. A caller who can sort it out themselves rings off satisfied rather than annoyed.
- Do not promise a wait you cannot keep
- Telling a caller their call will be answered within two minutes is a promise your phone system will break on your behalf. Thank them for waiting instead.
- Leave out anything with a date on it
- Christmas hours and a June promotion are wrong for most of the year. Keep the recording evergreen and make a second one when something genuinely changes.
One last thing: read it aloud before you buy. If you stumble over a sentence, the voice will too. Short sentences in plain words are written for the ear, and the ear is all your caller has. If you would rather someone else looked at it, send it to support@onholdmusic.ai before you pay.
Make your phone sound like your business
Start from a blank page or let the assistant draft it. Either way you hear five seconds of your own audio before you decide anything.
No account needed. Payment comes after the preview, not before.