The warning, with the invented precision taken out
This page carries a warning about representation agreements signed outside the union’s franchised system. It is the most consequential thing here and two of its details do not survive checking.
We told you such an agreement could carry a hundred per cent commission and run for seven years. We went back to the union’s own material for both figures and could not find either.
The claimWhat we could establish
A hundred per cent commissionOur own page
Unfound
No union source states it. A franchised agreement caps commission at ten per cent, and an agreement outside that system is not bound by the cap — but unbounded is not the same as a hundred, and we wrote the number as though somebody had reported it.
A seven-year termAlso ours
Unfound
Nothing ties that duration to these agreements. Union-approved arrangements typically start with a one-year term and renewal conditions. Seven appears to have come from nowhere.
It sits outside the union’s protectionsThe union’s own guidance
Verified
Their words: many such agreements “do not mirror the terms and conditions contained in the union franchise” and “may attempt to whittle away the overall union protections.” They advise getting independent legal advice before signing one.
And arbitration is the thing you loseThe mechanism
Verified
Dispute resolution through the union’s own department and then arbitration is a protection attached to franchised agents. Outside that system it does not apply, which is what “the union cannot help you” actually means.
We want to be exact about why this matters more than an ordinary correction. This is a warning. A reader deciding whether to sign something is entitled to numbers that came from somewhere, and two of ours did not. And the true version is not weaker. The danger is not a specific commission rate; it is that the ceiling and the appeal route both vanish, and that you find out which one you needed after the disagreement has started. That is worse than a number, because a number can be read and refused. An absent protection cannot be seen at all. If you have paperwork with an agency’s name on it, check it against the union’s franchised list today and take independent advice. That instruction is unchanged and it is the only part that was ever load-bearing.
Two hundred auditions, and the author who doubled her own figure
This page said the famous number cannot be traced to any primary document and that we had looked. That was right. What we did not have is the thing that proves it.
The figure does not travel alone. It moves in a block of four statistics, always in the same order, with the same attribution — and one of the four changes size depending on where you read it.
Where it appearsWhat it says
A trade magazine, October 2017Updated August 2019
900%
A nine hundred per cent increase in voiceovers required over the last three years. Two hundred auditions to book a job. Nearly nine thousand union voiceovers a month. Contenders narrowed to about ten per cent.
The same author’s own websiteCarrying a 2024 date
1800%
An eighteen hundred per cent increase. Same attribution, same three companions, same order. The lead figure has doubled and nothing else has moved.
And the source it is credited toBoth versions
No
Both say the figures come from “SAG-AFTRA and the US Department of Labor and Statistics.” There is no such agency. The name merges the Department of Labor with the Bureau of Labor Statistics, which are two different bodies.
So the demonstration is better than the absence, and it is more precise than we expected. This is not content quoting content until a detail drifts. It is one writer publishing nine hundred per cent in a trade magazine and eighteen hundred on her own site, under the same credit to an organisation that does not exist under that name. We went back to both possible primary sources and found nothing. The union’s own voiceover page carries rate sheets and contact details and not a single statistic. The federal labour handbook covers actors generally and contains no voiceover figure of any kind. One small correction to our own working: we were going to say the phrase “over the last three years” had survived unchanged for seven years. In the doubled version the author quietly changed it to “in recent years”, which is arguably the most honest edit in the whole trail.
And we are not going to hand you a replacement number, because the obvious one is no better. The screen-side booking ratios that circulate — one in twenty, one in a hundred, agencies expecting one in fifteen to twenty-five — do not trace to a single identifiable source either. They are a composite assembled across content sites. The one real trade candidate is a pseudonymous column from 2010 giving different figures entirely. Nobody knows how many auditions it takes, in either medium, and a coach who tells you with a straight face has not checked.
The best free thing on this page is not open to you
For a broke beginner the single most valuable line we published was to use a union foundation’s free voiceover lab. We re-checked its access rules. A British reader cannot get in.
It still exists and it is still genuinely free. What it is not is open. It is for members of the American performers’ union or the American stage union, or for people who have earned eligibility through American or Canadian union work.
How eligibility actually works.
Three days as a background actor on a union job, or one day in a covered principal role, or a year of paid-up membership in an affiliated North American union plus a principal engagement in its jurisdiction.
Then an application and a two-hour orientation.
British Equity is not one of the affiliated unions on that list. So the route we recommended is closed to most of the people we were recommending it to.
And when we went looking for the British equivalent of any of it, there is not one. There is no British member-vetted voice-over trade association — nothing that does what the American bodies do. The organisation whose name looks like a match is a clinical and academic charity for vocal health, run largely by ear nose and throat surgeons and speech therapists, and calling it a trade body would be simply wrong. There is no free institutional training route we could find — the conservatoire short courses are real and they are paid. And there is a legal wrinkle with no American parallel: British competition law prevents Equity from setting even a recommended rate for audiobook narration, so the one rate guide a British reader would most want cannot exist. The closest thing to a public British rate reference is published by a company that sells training courses.
We are stating that as a finding rather than hedging it. A British voice actor has a union, a clinical voice charity that is not about their career, no free institutional route, and a commercial training company doing much of the work an association would otherwise do. That is the landscape. Knowing it is worth more than being pointed at an American door you cannot open.
The document we recommended without reading
Our research proposed we point readers at a best-practice standard covering coaches, and admitted in the same breath that it had not read it and we should before recommending it. We read it.
It is real, it is free, and the coaching section is not a token third word in a title. It is a numbered set of standards written specifically about what a coach owes a student, and several of them land exactly on the problems this page already warns about.
On honest assessmentTo “honestly evaluate a student’s abilities in a fair and objective manner, without regard for monetary gain.”
The core
On the demoNever to rush a student into a professional demo “until both coach and student are reasonably convinced that the student is ready.” That is the demo mill, named by the industry itself.
The tell
On kickbacksNot to accept referral fees from demo producers, agents or casting directors unless full disclosure is made to the student.
The conflict
On acting as your agentTo refrain from it entirely, “understanding that this is an inherent conflict of interest.”
The door
So the fifth test for choosing a coach writes itself, and it is a document rather than a vibe. Ask whether they have read it. A coach who has voluntarily signed up to a third-party standard that tells them not to rush your demo and not to take kickbacks has made a small real commitment against their own commercial interest. Three things to be accurate about, though. The association is American, incorporated in Nevada, and while a British reader can join it carries no standing here. The membership figure we were given is stale — it was over five hundred more than a decade ago, not three hundred now. And the studio review is an internal panel of five members, not a panel of credentialed acousticians, which is a smaller claim than the one we were handed. The document is the useful part and it is genuinely useful.
What a coach can and cannot do
The case for coaching is mechanical rather than mystical, and the case against most of what is sold alongside it is that nobody selling it owns the door.
You cannot hear yourself. Your voice reaches your own ears through the bone of your skull as well as through the air, so the sound in your head has never once been the sound in the room. That is the entire justification and it is enough.
What they can do
Be the outside ear, direct you while you record, and tell you when you are not ready — which is the most valuable and least marketable sentence in the business.
What casting actually wants
Actors, not voices. Named animation casting directors at three major studios say versions of the same thing — one of them plainly: “We want people’s natural voices... the standard demos don’t really work for us.”
What they cannot do
Get you an agent, or open animation, games and anime. Those run on agent submissions and closed lists. Anyone promising representation is either mistaken or has an interest they have not declared.
One correction to our own sourcingWe had those three casting directors down as saying one thing together. They are all real, credentialled and on record, but we could not find a single joint quote. We are attributing each of them separately from now on.
One claim we are softening. We described a particular anime studio’s casting as running through a closed database, and we could not verify that from the company’s own material. The general point about that corner of the industry holds and the specific attribution does not, so we have taken the company’s name off it.
It took the ladder
The honest answer to whether this is worth starting now is the hardest section on the page, and it survives checking with one refinement.
Synthetic voice has not replaced the top of this industry. It has replaced the work people used to learn on — the small, unglamorous, badly paid jobs that were how anybody got competent enough to be worth hiring for the good ones.
The survey behind that argument is real and one of the two figures we quoted is not confirmed. The association’s 2026 survey ran in January and February and drew nearly fourteen hundred responses. Twenty-one per cent of voice actors reported losing work directly to synthetic voice, up from fourteen per cent the year before. That figure we can stand up. The second figure we published, of thirty per cent, we could not locate anywhere in the survey’s own published sections, so it is off the page until somebody can point at it. And the observation we made about the survey is worth naming as a principle rather than an aside: it can only see people who are already working. The beginners who never booked anything are in nobody’s data at all, which is the same blind spot that makes every survey of working actors’ incomes look better than the profession does.
What follows from it is uncomfortable and it is not despair. If the ladder is gone, the only route left is arriving already able — which makes the outside ear more valuable rather than less, and makes buying a demo before you can act even more clearly a waste of money.
How to choose one
Four tests, and a fifth that we can now add because it points at a document.
Do they actually work?
The testRecent, checkable credits in the thing they are teaching.
WhyTeaching a corner of the industry you have never worked in is specifically covered by the standards document, which says it must be disclosed.
Do they teach acting?
The testWhether the session is about performance or about your voice as an object.
WhyEvery casting director on record says they are hiring actors. A beautiful voice with nothing behind it books nothing.
Is the coaching separate from the demo?
The testWhether a demo is scheduled in advance or earned.
WhyA demo made before you can act is not a credential, it is a receipt. And the standards document tells coaches not to rush it.
Do you record, and get directed?
The testWhether you leave with audio and notes on it.
WhyThe whole product is the outside ear. A session where nobody redirected you has not delivered it.
Have they read the standards?
The testAsk. It is one question and it takes thirty seconds.
WhyIt is checkable, it is free, and it is the only one of these five that is a document rather than a judgement.
If you have already paid
Somebody reading this has spent money on something that did not work. This page has spent a lot of words on warnings and owes that reader something else.
A bad course is a sunk cost, not a verdict. What you needed was an outside ear. If you did not get one, you still have not had one, and nothing has been lost except money.
Three things worth doing this week.
Do not send the demo. If you have one you were not ready for, sitting on it costs nothing, and a bad demo spends an introduction you only get once.
Check any paperwork with an agency’s name on it against the union’s franchised list, and take independent advice if it is not there. Per the section at the top of this page, that is the part with real consequences.
And go and find the free things. Nobody administering a free programme asks what you spent last year.
Grading the sources
A page about people quoting numbers they have not checked turned out to be quoting two numbers it had not checked.
The claim, and where it came fromWhat we found
A hundred per cent commission, and a seven-year termOur own warning
Unfound
Neither figure appears in any union source we could reach. The structural warning around them is verified and did not need them. Invented precision inside a safety warning is the worst place for it.
The free voiceover lab is the answer for a beginnerOur own recommendation
Not here
Still free, still excellent, and gated behind American or Canadian union credentials. British Equity is not an affiliated union. Most of our readers cannot use it.
A thirty per cent figure from the 2026 surveyOur own page
Unfound
We could not locate it in the survey’s own published sections. Removed until somebody can point at the question it answers.
Three casting directors saying one thingPresented jointly
Separate
All three are real and credentialled. There is no joint quote. The panel it appears to come from was members-only with no recording permitted. Attributed individually now.
The lead statistic doubles between publicationsThe two pages themselves
Primary
Nine hundred per cent in the trade magazine, eighteen hundred on the same author’s own site, same attribution, same companions. Quoted verbatim from both.
The agency credited does not existBoth versions
Primary
The name merges two real bodies. A figure credited to an organisation nobody could look up has not been checked by anybody who repeated it.
There is no primary sourceChecked at both
Null
The union’s voiceover page carries no statistics at all. The federal handbook has no voiceover figures of any kind. Our original null result holds.
The standards document covers coaches properlyRead at last
Primary
Ten numbered standards, several of them directly about demos, kickbacks and acting as an agent. Better than the recommendation we were given, which had not opened it.
The 2026 survey and the twenty-one per centThe association’s own publication
Good
Nearly fourteen hundred responses, run in early 2026. Twenty-one per cent losing work to synthetic voice, up from fourteen. Dated and current.
A named studio casts from a closed databaseOur own page
Softened
Plausible and consistent with how that corner works, and not confirmed by the company’s own material. The name is off the claim.
Three hundred members, acoustics expertsGiven to us
Overstated
Over five hundred more than a decade ago, so the figure is stale rather than wrong. And the studio review is a five-member internal panel, not credentialed acousticians.
Screen-side booking ratios as a substituteSuggested to us
Also folklore
The ranges do not trace to any identifiable source. They are a composite across content sites. We are not swapping one unsourced number for another.
One live example worth naming, because it is the failure with a methodology statement bolted on. An aggregator publishing voice-acting statistics states that figures failing independent corroboration are excluded regardless of how widely cited — and publishes untraceable numbers on the same page, under a caption admitting that attribution is at report level rather than per statistic. Its own footer contradicts its own promise.
Where we differ from the standard account
How many auditions to book one?
Commonly saidTwo hundred, credited to the union and a federal agency.
What we foundThe agency does not exist and the companion figure doubles between the author’s two publications.
Is there a standard for coaches?
Commonly saidNo accreditation, so nothing.
What we foundNo accreditation, but a free published best-practice document written specifically about coaching.
Where does a broke beginner start?
Commonly saidThe union foundation’s free lab.
What we foundExcellent, and gated behind American union credentials. British readers cannot get in.
Does Britain have an equivalent?
Commonly saidAssumed, or never asked.
What we foundNo trade association, no free institutional route, and a legal bar on the rate guide people most want.
What does a bad agreement cost you?
Commonly saidSpecific frightening numbers, including ours.
What we foundThe commission cap and the appeal route both disappear. That is worse and it is what the union actually documents.
Do you need a demo?
Commonly saidEvery package ends in one.
What we foundThe industry’s own standards tell coaches not to rush you into one. It is earned, not scheduled.
Is it worth starting, with synthetic voice?
Commonly saidEither doom or denial.
What we foundIt took the ladder rather than the top, which makes arriving able the only route left.
Do one thing tonight. If you have signed anything with an agency’s name on it, go and check that name against the union’s franchised list. It takes two minutes and it is the only item on this page with consequences you cannot undo later. Everything else here is money. That one is your representation.
