Here is a situation I have seen more than once. A client books a 30-second read for what the email calls a “digital campaign.” The recording is delivered on Tuesday, everyone is happy, the invoice is paid. Six months later the same ad turns up on television during a cricket match.
Nobody in that story was being dishonest. The brief simply never said where the ad would run or for how long, and so the quote was written for one thing while the campaign quietly became another. That is not a billing dispute. It is a scoping failure, and it is entirely preventable. Here’s how:
Three words prevent it: usage, term, and territory. They are the levers that decide what a voice over costs and, more importantly, what you are allowed to do with it. I have written about the cost side already. This post is about what you are actually buying.
The recording is the small part
When you hire a voice artist, you are paying for two separate things that arrive in the same delivery.
The first is the work: the session, the read, the direction, the editing, the mastering, the file that lands in your inbox. That part is roughly fixed. A 30-second script takes about the same effort to record well regardless of what happens to it afterwards.
The second is the licence: your permission to use that recording in specific ways, for a specific time, in specific places. That part is not fixed at all. It is the part that scales, and on any commercial project it is usually the larger half of the number.
This is why “but it is only 30 seconds” does not move a quote much. The length tells me how long the session runs. It tells me almost nothing about what the recording is worth to you. It also tells me nothing about how you will be using my voice.
Why this is a licence, and not a sale
There is a part of this that is easy to miss, and I would rather say it plainly than leave it implied.
Start with the part that is not special. Licensing by usage, term, and territory is simply how most creative work is bought. Photographers do it. So do illustrators, composers, and type designers. If you have ever paid for a rights-managed image, or cleared a sync licence for a piece of music, you have already run this exact negotiation. None of it is a voice-over peculiarity, and none of it is me inventing friction.
What is different here is not the commercial model. It is what sits underneath it. A photographer licenses an image they made, and the image is a thing they can hand over and step away from. I am licensing my voice, which is less something I produced than something I am. When your ad runs, that voice is in someone’s car, kitchen, or headphones, working for your brand, and it arrives attached to a person, because there is no version of it that is not me.
So what you are licensing is not really thirty seconds of audio. It is a piece of someone’s identity, put to work for a commercial purpose, for a while. That is a completely ordinary thing to buy, and I am glad to sell it. It is just worth stating what it is, clearly.
This is also where the fee comes from, and why it scales the way it does. Wider usage means more of that identity in circulation, attached to your brand rather than to anyone else’s. A longer term means longer before I can decide otherwise. A bigger territory means more places where people hear my voice and think of you.
There is a quieter cost sitting inside that. Once my voice is working for you, I can’t work for other brands in the same category, whether you say so or not, because nobody wants someone else’s brand attached to theirs. That constraint costs me work, it lasts as long as your term does, and it is already inside the number you were quoted. If you want it guaranteed rather than merely likely, ask for exclusivity by name and I will quote it separately, with the category defined and an end date on it.
None of that is a penalty, and none of it is precious. It is the honest shape of what is being licensed.
It is why the arrangement is a licence that expires rather than a sale that is final. It is also why the AI restriction further down is absolute rather than a matter of price. You are borrowing something I cannot replace.
Usage: what the audio is for
Usage is the single biggest lever, and it is the one most briefs leave out. Ranked roughly from lowest to highest value:
- Internal, non-broadcast. Staff training, onboarding, induction, internal comms, an e-learning module behind a company login. The audience is your own people, and it is a finite number.
- Corporate and explainer. A product walkthrough on your own website, a video on your YouTube channel, a demo reel for a trade stand. Public, but not advertising.
- Organic social. Posted to your own channels, unpromoted. Reach is whatever the algorithm gives you.
- Paid digital. The moment money goes behind a post, it becomes advertising, and it is priced as advertising. This is the line most people cross without noticing.
- Broadcast. Television and radio. Highest reach, highest production value, highest rate, and usually the shortest terms.
- Out of home and in-store. Airports, malls, cinemas, retail audio, transport announcements. Priced by scale of installation, not by seconds.
- IVR and on-hold. A special case. Low glamour, long life, usually licensed generously because it runs quietly for years.
Two projects with identical scripts and identical runtimes can sit at either end of that list. The read is the same. The product is not.
Term: how long you get to run it
Term is how long your licence lasts, measured from first air date rather than from delivery. The common shapes:
- Three or six months. A seasonal push, a product launch, a festival campaign.
- One year. The default for most digital and broadcast advertising, and the assumption behind most rate guides.
- Two or three years. Longer campaigns, brand platform work, anything you expect to keep running.
- Perpetual. No end date. Genuinely useful for corporate explainers, e-learning, and IVR, where re-recording every year would be daft.
A longer term is not linear. Going from one year to three does not triple the price, because the second and third years are worth less to you than the first. But it does move the number, and it should, because it removes my ability to license that read elsewhere and removes your need to ever come back.
One practical note. If a campaign is shelved and then revived eighteen months later, the term has usually run out even though the ad barely aired. Say so at the start and I can write the term to begin at first air date instead of at delivery. It costs you nothing to ask.
Territory: where it runs
Territory is geography, and it stacks on top of the other two:
- Single city or state. Regional retail, local services, a state-level government campaign.
- National. Pan-India.
- Regional cluster. India plus the Gulf, or South Asia, which is very common for brands with a diaspora audience.
- Global. Anywhere, no restriction. This is what most overseas clients hiring an Indian English voice from abroad actually need, and it is worth naming explicitly rather than assuming.
Territory catches people out with digital work, because the internet does not respect borders. If the ad is geo-targeted to three cities, that is a regional buy. If it is on an unrestricted YouTube channel that anyone can reach, that is global, whatever the media plan says. Be honest about which one you are actually running, because the platform’s targeting settings are the evidence either way.
The five words that start most arguments
Buyout. The worst offender, because it means completely different things to different people. To some it means “all rights, forever, everywhere.” To others it means “one year, all media, national.” It has no fixed definition in the Indian market. If it appears in a purchase order, define it in the same sentence or expect trouble.
Full rights. Same problem, different clothes. Full rights to do what, where, until when?
Unlimited. Usually means unlimited platforms, not unlimited time or territory. Almost never means all three.
In perpetuity. This one is precise, and it is fine. It means no end date. It does not automatically mean all media or all territories, so it can sit quite happily alongside a narrow usage: perpetual, internal training, India only, is a perfectly sensible and affordable licence.
Work for hire. Imported from American contracts, where it has specific legal weight around authorship. In India it is often used loosely to mean “we own it.” If you need actual assignment of rights rather than a licence, that is a real and buyable thing, but it needs to be said plainly and priced as what it is.
None of these words are traps. They are just imprecise. Replacing any of them with a sentence naming usage, term, and territory removes the ambiguity completely.
Same read, three licences
Take one 30-second script, recorded once, in one session. Nothing about the audio changes.
- Internal induction video, perpetual, India. The cheapest of the three by a wide margin. Finite audience, no advertising value, no reason to ever revisit it.
- Paid social and YouTube pre-roll, one year, India. Several times the first. It is advertising, it has budget behind it, and the licence expires.
- Television and all digital, two years, global. Several times again. Broadcast reach, long term, no geographic ceiling.
The gap between the top and the bottom of that list is large. If you have ever wondered why an artist cannot give you a rate card with one number on it, this is why. The number is not a fee for reading. It is a price for a specific set of permissions, and I do not know which permissions you need until you tell me.
What happens when the term ends
The ad comes down, or you renew. Both are normal.
Renewals are usually the easiest conversation of the whole project. The recording exists, the relationship exists, and the price for a second year is often lower than the first. Diarise the expiry date when the campaign launches, and it becomes a two-email formality instead of a scramble.
What is genuinely a problem is the third option: leaving it running and hoping nobody notices. Voice work does get noticed, often by other artists, and the fix after the fact is always more expensive and more awkward than the renewal would have been. I would much rather send you a renewal invoice than a letter.
The AI clause, and why it is now standard
This is new enough that a lot of purchase orders have not caught up, so I will be blunt about it.
No standard voice over licence grants the right to use a recording to train a machine learning model, to build a synthetic or cloned voice, or to generate new speech in that voice. Those are separate rights, and in my case they are not merely unpriced. My service terms exclude them permanently, at any price, and the restriction survives delivery and payment.
This matters to you and not only to me. If a downstream vendor in your chain feeds delivered audio into a voice cloning tool to patch a line, that is unlicensed use happening under your brand’s name, and it lands on your desk before it lands on mine. Ask the question early, put the exclusion in writing, and the whole problem disappears. I have written more about where synthetic voices actually sit right now.
Put it in the brief
You do not need legal drafting for this. One line in the enquiry email does the job:
Corporate explainer, 480 words, paid digital and organic social, one year from first air, India and the Gulf. No broadcast. Delivery by the 28th.
That is a complete brief for pricing purposes. Everything I need to send back one accurate number the same day, instead of three rounds of clarifying email. If you want the fuller version of what makes a brief useful, I have written that up separately.
If you genuinely do not know yet, say that too. “Digital for now, television possible in Q4” is an honest brief, and I will quote the digital licence and price the broadcast upgrade alongside it, so the number exists before you need it. Uncertainty is fine. Silence is what causes the problem.
When you genuinely do need everything
Sometimes you really do want perpetual, global, all media, no restrictions. A brand voice you will reuse for years. A product that ships worldwide. An IVR system nobody wants to re-record.
That is a completely reasonable thing to buy, and I will quote it happily. It costs what it costs, because it is the largest version of the product and it removes every future conversation at once. Plenty of clients decide it is worth paying once and never thinking about it again. The only bad outcome is buying the small licence, using it like the large one, and discovering the difference in year two.
My standing position on all of this, including what happens when a brief specifies no scope at all, is set out in my service terms. The short version: with no scope specified, the licence covers the single use described in the brief, and nothing wider. Naming usage, term, and territory is how you get something better than that.
Working out what you need? Send me the script or the word count, where it runs, for how long, and on what platforms. That is all it takes to get a clear number back. Get in touch here or call +91-99620-28747.