How to Make a Voice Over Demo Reel

Somewhere in your search history there is a page telling you that the demo is the beginning. Get a demo, it says, and the work follows. That sentence has sold more coaching packages than any other sentence in this business, and it is backwards.

A demo is not a portfolio. A portfolio invites someone to browse. A demo does the opposite: it is handed to a person who is behind on their day, who has forty other files in the folder, and whose actual job in that moment is to reduce forty to four. They are not listening for a reason to hire you. They are listening for a reason to stop.

Understand that, and everything else about a demo — its length, its order, its first breath — stops being a matter of taste and becomes a matter of engineering.

What a demo is actually for

A demo answers exactly one question: can this person do the thing, at a professional standard, today? Not could they, with training. Not are they interesting. Can they do it now, for money, without the client having to explain anything twice.

It is a screening tool, and screening tools are unfair by design. That is the point of them. Nobody at an agency is going to give your file three chances, because the file is not the actor — it is a proxy for the actor, and a proxy only has to be plausible enough to survive the cut.

And here is the part nobody says out loud: a demo does not buy you access. On Voices.com, the free Guest tier lets you upload unlimited demos and does not let you reply to a single job. You can be privately invited; you cannot audition. To audition you pay $499 a year. So a demo, on its own, on the biggest platform in the world, is a file sitting in a room nobody has to walk into. Access is bought separately. The demo just decides what happens if someone opens the door.

Length and structure — what the file specs quietly tell you

Nobody publishes an official demo length, and anyone who tells you the industry standard is exactly ninety seconds is repeating something they heard. But the platforms do publish their upload specs, and the specs give the game away.

Voices.com wants an MP3 at 44.1kHz, 128kbps, and between one and two megabytes. Do the arithmetic on that: 128 kilobits a second is about one megabyte a minute. Their file ceiling is, quite literally, about two minutes of audio. Voice123 asks for MP3 at 44.1kHz, 96 to 128kbps CBR, 16-bit, mono. Backstage asks for 44.1kHz, 16-bit mono MP3, at 128kbps or less. The industry did not agree on a demo length in a meeting. It agreed on an upload limit, and the upload limit decided for everyone.

So: a demo is a small file, at modest bitrate, in mono. Which means the reel that lives in your head — the sweeping montage with the music beds and the twelve characters — does not fit inside the box the business actually uses. Build for the box.

On structure, the honest guidance is short. One demo per category. A commercial demo is a commercial demo; an animation demo is a separate file. Do not mix them, because a listener screening for commercial reads will hear your dragon and conclude, fairly, that you did not understand the brief. And within a demo, every segment should sound like a different job, not a different voice. That distinction is the whole chapter, and I will come back to it.

The first few seconds decide everything

Whatever is at the top of that file is your demo. The rest of it is optional listening, and most of it will not be listened to.

So the first spot is not your favourite spot. It is not the one you found hardest, or the one your coach praised, or the one where you do the accent. It is the spot that sounds most like the work you want to be hired for tomorrow, delivered at the level of the person who currently gets hired for it. If your strongest ten seconds are at 0:40, you do not have a demo. You have a demo with forty seconds of throat-clearing glued to the front of it.

And it has to sound like you. This is where beginners are most badly served by the advice they are given. Jamie Sparer Roberts, who casts for Walt Disney Animation Studios, has said plainly that most people think animation casting is about finding people who can put on voices, and that she almost never hires someone to do a voice that is not their own, or a slight variation of it. The animation casting director Mary Hidalgo puts it even more bluntly: you are never going to do cartoons if you do not know how to act. A demo full of impressions is a demo that answers a question nobody asked.

The demo mill

Chapter Five names the demo mill as a warning sign, and this is the chapter where it does its damage, so let us be specific.

A demo mill is a coaching outfit whose actual business is not teaching. It is selling you a demo. The classes are real enough, the coach may even be good, but the structure of the thing bends everything toward one outcome: a package, a studio day, a produced reel, and an invoice. You will be told you are ready. You will be told the demo is the thing standing between you and the work. Both of those statements exist to close a sale.

A demo made before you can act is money set on fire. Not wasted — burned, because a bad demo is worse than no demo. It gets you screened out by exactly the people you most want to reach, and it teaches you nothing about why. You cannot iterate on it. You paid to be rejected faster.

When you are actually ready

Here is a concrete test, and it is deliberately hard.

You are ready when you can take a piece of copy you have never seen, and a direction you did not expect — less warm, more urgent, take the smile out, land the last line harder — and change the read on the second take in a way the person directing you can actually hear. That is the skill being screened for. Not tone. Not a nice voice. The ability to receive a note and act on it. Every casting director quoted in the research says a version of the same thing: they are looking for actors, and actors who have theatre, improv and sketch training show up better in the booth than people who do funny voices.

You are also ready only when the room is. The demo is an audio product, and Backstage's own guidance is stark: the room you record in matters more than the mic, the interface, the preamps, the computer and the software. If you have $1,000 for a mic and nothing for acoustics, split it in half. An untreated bedroom typically sits at a noise floor of -40 to -50 dBFS, and the professional bar is -60dB or lower. A great performance in a boxy room is not a demo. It is a good take you cannot use.

Trade commentary describes being agent-ready as three things: a website, a calibrated home studio, and a demo that does not embarrass you. Notice the order. Notice that the demo is last on that list too.

What it costs — and why nobody will tell you

I am not going to give you a price, because I could not find one I could stand behind. Demo production is priced privately, quoted per client, and bundled inside coaching packages precisely so that it cannot be compared. The published rate guides that the industry does use — GVAA, Gravy For The Brain — could not be fully verified in the research behind this chapter, and neither of them prices demo production anyway. They price the work, not the shop window.

The one demo-related figure I can source is SAG-AFTRA's, and it is for a different thing entirely: under the 2025 Audio Commercials Contract, a session for demos, copy tests and non-air material is $251.53 for an hour. That is an agency paying you to record a test version of a client's ad. It is a category of paid work. It is not the reel with your name on it, and if anyone quotes it at you as though it were, you have learned something useful about them.

So the price of a demo is opaque, and the opacity is not an accident. When a market hides its prices, the reason is almost always that the prices would not survive comparison. Get three written quotes. Ask what you are actually buying — studio hours, script writing, direction, editing, music licensing, mixing — as separate line items. And ask what happens if, on the day, the producer thinks you are not ready. The answer to that last question tells you what kind of business you are standing in.

One last thing, and it is the hard one. The demo used to be the first rung of a ladder: you made one, you got the small corporate jobs, the phone systems, the cheap e-learning, and you climbed. Those are the exact categories that synthetic voice is eating first. The top of this industry is largely holding. The bottom is not. Which means a demo, in 2026, is no longer a ticket to the bottom rung — it is a claim that you belong further up, and it had better be true when someone presses play.