Commercial Voice Over: A Buyer's Guide

Commercial voice over for brands: script timing for 15, 30 and 60 second ads, casting, usage-based rates, AI voice consent and broadcast loudness specs.

Published 2026-10-25 · Video Marketing · Neverframe Team

Commercial Voice Over: What Brands Need to Know Before Booking a Voice

Since December 13, 2012, every television commercial aired in the United States has had to match the average loudness of the program around it. That rule, set by the FCC under the CALM Act, is one of the few hard technical lines in commercial voice over, and most brand teams have never heard of it. Everything else depends on choices the brand team makes about the voice, the read, the rights and the delivery files.

This guide is written for the people who commission ads. If you are a marketing director, brand manager or agency producer planning a spot with narration, you will find the decisions in the order you will meet them: when a commercial voice over is the right tool, how to write a script that fits 15, 30 or 60 seconds, how to cast and direct, how rates are structured, what changes when AI voices enter the picture, and what a finished audio package should contain.

In brief, a commercial voice over is a recorded narration made for paid media, priced on usage (where it airs, for how long, on which platforms) rather than on the hours spent in the booth. Lock the script before you cast, cast against the brand and the audience rather than a demo reel, put usage terms in writing before the session, and ask for a mix that meets broadcast loudness specs even if the ad starts on social.

What Counts as a Commercial Voice Over

The term covers any voice that speaks to the audience from off screen in a piece of paid advertising: TV and streaming spots, pre-roll on YouTube, paid social, radio and podcast ads, cinema ads and retail screens. The voice may narrate the whole spot, close it with a tagline, or carry the legal copy at the end.

The craft has a long history in broadcasting. Wikipedia's entry on voice-over describes how the format grew with the medium:

"The commercial use of voice-over in television advertising has been popular since the beginning of radio broadcasting. In the early years, before effective sound recording and mixing, announcements were produced "live" and at once in a studio with the entire cast, crew and, usually, orchestra."

The same entry describes how the 1960s and 1970s turned the voice into a selling tool, a shift that still shapes how ads are cast:

"The 1960s and 1970s witnessed a significant increase in the use of voice-overs in commercials. Advertisers began to realize the impact of a strong, persuasive voice in selling products."

What separates a commercial read from other narration is the commercial context, and this is where buyers usually get confused. The same actor reading the same 75 words can be priced very differently depending on whether the recording lives on an internal training portal or runs as a paid ad across three countries. Voices.com, one of the largest voice marketplaces, draws the line on its rate page, describing broadcast work as "anything with paid promotion behind it, TV, radio or online ads." Pricing for that work follows the media plan, and the same page spells out what sits on the other side:

"These are projects that are not tied to advertising, think training videos, explainer content, podcasts, or internal communications."

So a narrated product video on your website is usually non-broadcast. The moment you put media spend behind that same file, it becomes a commercial and the usage terms change.

Voice over, announcer, spokesperson: the differences that matter

Brands often use these words loosely. They lead to different bookings, so it helps to keep them apart.

A recognizable voice becomes a brand asset over time, and an asset like that needs a contract that matches how long and how widely the brand plans to use it.

If your ad needs a face as well as a voice, you are in spokesperson territory, which brings in shoot logistics and a talent release. Our guide to the talent release form for video production covers the paperwork for people who appear on camera.

When a Voice Over Is the Right Choice for an Ad

Plenty of strong spots run on music, sound design and on-screen type alone. A voice over earns its place when one of these conditions is true:

• The product needs explaining in the time available, and text on screen would be too dense to read.
• The ad will run where people listen more than they watch: connected TV with the phone in hand, audio platforms, pre-roll that plays while the viewer looks elsewhere.
• The brand already has a recognizable voice and the spot should sound like the rest of the campaign.
• The spot has an offer, a deadline or legal copy that must be heard clearly.
• You will version the ad across markets or audiences, and a voice track is cheaper to swap than footage.

It is the wrong choice when the visuals carry the story better on their own, when most of the media is silent autoplay feeds, or when the script is being used to cover for footage that does not work. Many social feeds start videos muted, so the narration must be backed by captions and strong supers. Our social media video specs guide lists the formats you will need for those cuts.

A quick decision checklist

Before you brief a voice, answer these questions in writing. If you cannot answer three of them, the spot is not ready for casting yet.

Writing a Commercial Voice Over Script That Fits the Time

The most common problem in a commercial voice over session is a script that does not fit. A 30 second spot leaves less than 30 seconds of speaking room once you allow for a beat of music at the open, the logo at the close and breathing space between thoughts. Writers who time their copy by reading it silently almost always overrun.

The Wikipedia entry on words per minute summarizes research on comfortable speech rates and puts the recommended pace for audiobooks at 150 to 160 words per minute, with slide presentations slower, at 100 to 125. Ads are often read faster than audiobooks, but the audiobook range is a sound ceiling for copy that has to be understood on first listen. At 150 words per minute, a full 30 seconds holds about 75 words. Take off the open and the close, and a realistic target for a 30 second spot is closer to 55 to 65 words. A 15 second cut holds roughly half of that, which means a different script, not the 30 second script read faster.

These are planning numbers derived from the speech rates above. A fast, energetic read can carry more; a warm, slow brand read carries less. Time the script out loud with a stopwatch, at the pace you want, before anyone is booked.

Script habits that save session time

• Write for the ear. Short sentences, one idea each, and numbers written the way they should be spoken (twenty-nine ninety-nine for $29.99).
• Mark the words that must land. Underline the product name and the call to action so the director and the talent agree on emphasis.
• Give pronunciation for every brand name, product name and place name, with a reference recording if the name is unusual.
• Separate legal copy. Disclaimers are often recorded as a separate track at a different pace, so they can be swapped without re-recording the whole spot.
• Lock the 30 and the 15 together. If the 15 is not a clean edit of the 30, write it as its own script.

For the broader craft of commercial writing, our video script writing guide walks through structure, hooks and revisions.

Casting the Voice

Casting is where most of the creative value is won or lost. The voice tells the listener who the brand is before the words do. A voice that is wrong for the brand cannot be rescued later by music or EQ.

Start with a casting brief rather than a list of famous examples. A good brief describes the listener, the attitude of the brand, the energy of the read and the role the voice plays. Warm, confident, mid-thirties, sounds like a friend who knows the product, never sells is useful. Like the voice in that car ad is not, because nobody agrees on which car ad.

Where brands find commercial voice talent

Marketplaces have made auditions cheap. Voices.com states on its rate page that "Finding and Auditioning Talent Is Always Free" and that "You only pay once you've chosen the right voice for your project." The talent fee remains the main cost, and the page describes it plainly:

"This amount is set by the talent and may vary based on factors like project scope, usage, and their level of experience."

A shortlist of five to eight auditions, all reading the same 15 to 20 seconds of the actual script, is usually enough. Listen on the devices your audience will use: phone speakers, a laptop, a TV. A voice that sounds rich on studio monitors can disappear on a phone.

Union or non-union

In the United States, much broadcast commercial work is covered by SAG-AFTRA contracts, which set minimum terms, session fees and residual structures for union members. Non-union talent negotiates terms directly. Neither route is better in itself. Union work brings predictable terms and access to talent that only works under contract; non-union work can be simpler for regional or digital-only campaigns. If you work with an agency or a studio, ask which contracts they are set up to administer before you fall in love with a voice you cannot book.

Directing the Session

Most commercial sessions today are remote. The talent records in a treated home or professional studio, and the client listens in over a live connection. Remote sessions are usually booked in short blocks, so the materials need to be ready before the call starts.

Before the session, send the final script, the pronunciation guide, a rough cut or animatic with temporary audio, and two or three reference reads that capture the tone. During the session, keep one voice giving direction. Several people speaking to the talent at once produces conflicting notes and tired reads.

Useful direction describes intent rather than mechanics. Say it like you are letting a friend in on a secret gets a better take than softer on the third word. Ask for a few alternate reads of the key lines: a warmer one, a more direct one, one with a smile in it. Those alternates are what save you in the edit.

Record wild lines as well: the brand name alone, the call to action in three tones, the tagline at two speeds. They cost seconds in the session and solve problems weeks later when a cut changes.

How Commercial Voice Over Rates Work

Buyers often ask for a single price. There is no single price, because commercial voice over is priced on how the recording will be used. Voices.com explains the logic on its rate page, where it says that "Pricing here is based on usage, including: Location: where it's airing Duration: how long it's running Platform: TV, radio or online." It also gives the reason for the premium on paid media:

"Because these ads reach larger audiences, they come with higher rates to fairly compensate the talent for their voice's exposure and impact."

Its FAQ adds that "Factors like where the voice will be heard, how it will be used, and for how long all impact the cost. The broader the reach, the higher the rate." For non-broadcast work the same source describes a simpler model, noting that such jobs are "typically priced per word or per finished minute." On top of the talent fee, the platform adds a small processing fee that depends on the payment method.

The Global Voice Acting Academy publishes a widely used rate guide for the voice over community. It is careful to present those numbers as observed market ranges rather than fixed prices, stating that the guide "reflects ranges observed in the market and is not an attempt to set, fix, or standardize industry pricing." Treat any published range as a starting point for a conversation with the talent or their agent.

The variables that move the fee

Ask for renewal pricing in the first contract, so year two is not a fresh negotiation from a weak position. And list every cut you expect to make (60, 30, 15, 6, tags, legal versions) at booking time, because adding cuts later is a usage change.

For a view of how voice fits into the whole cost of an ad, see our breakdown of how much a 30 second commercial costs and the video production budget guide.

Usage Rights and Contracts

The usage terms in a voice contract should mirror your media plan. If they do not, one of two things happens: you overpay for rights you never use, or you run the ad somewhere the contract does not cover and face a claim.

At a minimum, the agreement should state:

• The spots and cuts covered, by title and length.
• The media: TV, streaming, online paid, social paid, radio, audio streaming, cinema, out of home.
• The territory and the term, with start and end dates.
• Exclusivity, if any, and the product category it covers.
• Whether the recording can be edited, re-cut or combined with other material.
• Whether the voice can be used to create a synthetic or AI version, and on what terms.
• Renewal pricing and the process to extend.
• Credit or no-credit, and approval rights if the talent has them.

Music and voice are often licensed in parallel, and both have term and territory clauses that must line up with the media plan. Our guide to sync licenses for music in ads covers the music side, and the video usage rights guide covers footage and talent rights together.

AI Voices in Commercials: Consent, Law and Contracts

Synthetic voices have become good enough for real advertising. They are useful for fast versioning, for testing scripts before a session, and for internal drafts. They also bring legal exposure that did not exist a few years ago, mainly around using someone's voice without permission.

Tennessee was the first state to put voice explicitly into its likeness law. The governor's office called it "first-of-its-kind legislation updating Tennessee's Protection of Personal Rights law" and explained the gap it was meant to close:

"While Tennessee's preexisting law protected name, image, and likeness, it did not specifically address new, personalized generative AI cloning models and services that enable human impersonation and allow users to make unauthorized fake works in the image and voice of others."

The release then summarized the change: the act "builds upon existing state rule protecting against the unauthorized use of someone's likeness by adding" voice to what the law protects. Governor Bill Lee tied the law to the state's creative economy: "As the technology landscape evolves with artificial intelligence, I thank the General Assembly for its partnership in creating legal protection for our best-in-class artists and songwriters." The recording industry welcomed it too, with RIAA chairman Mitch Glazier saying "We applaud Tennessee's swift and thoughtful bipartisan leadership against unconsented AI deepfakes and voice clones." House Majority Leader William Lamberth framed the balance the law tries to strike:

"While we support the responsible advancement of this technology, we must ensure we do not threaten the future livelihood of an entire industry. This legislation is an important step in maintaining public trust and advancing ongoing efforts to protect and inform Tennessee consumers."

At the federal level, the U.S. Copyright Office has called for new protection. Its July 2024 release on digital replicas opens with the scope of the problem:

"This Part of the Report responds to the proliferation of videos, images, or audio recordings that have been digitally created or manipulated to realistically but falsely depict an individual. Given the gaps in existing legal protections, the Office recommends that Congress enact a new federal law that protects all individuals from the knowing distribution of unauthorized digital replicas."

The report is being published in parts, and the release lists what comes next: "Forthcoming Parts will address the copyrightability of materials created in whole or in part by generative AI, the legal implications of training AI models on copyrighted works, licensing considerations, and the allocation of any potential liability." The Office said it "also offers recommendations on the elements to be included in crafting such a law." Shira Perlmutter, Register of Copyrights, added:

"I am pleased to begin sharing the results of our comprehensive study of AI and copyright, with this first set of recommendations to Congress."

Perlmutter framed the urgency in the same statement: "We believe there is an urgent need for effective nationwide protection against the harms that can be caused to reputations and livelihoods."

For a brand team booking a voice today, that legal picture shapes the brief and the contract:

• Never ask for a voice that sounds like a specific celebrity or actor, human or synthetic. Sound-alike requests are exactly what these laws target.
• If you use a synthetic voice, use a licensed stock voice or a clone created with the written consent of the person, with a contract that defines where and for how long the clone may be used.
• Put AI terms in every human voice contract too. State clearly whether the recordings may be used to train or generate a synthetic version.
• Keep records of consent and licenses with the campaign files, the same way you keep music licenses.

Union contracts in the US have added their own terms on digital replicas, so check the current agreement with your agency or the talent's representative before using any synthetic version of a union performer.

If your plan is to put a company leader on screen through a digital double, the consent logic is the same. Our AI avatar video production guide explains how that works for executive messages.

Recording, Mixing and Delivery Specs

Recording quality matters before the mix does. The Wikipedia entry on voice-over notes that the sound recording industry uses "presence" as its benchmark, and that "The term "presence" measures the legitimacy of how a voice sounds, specifically one of a voice-over." A professional booth, a good microphone and a quiet room are what give a read that presence.

Delivery is where an approved read most often runs into trouble. Broadcast and streaming platforms have loudness requirements, and a mix that ignores them will be rejected, turned down by the platform's normalization, or simply sound wrong next to the program.

In the United States, the FCC page on loud commercials explains the rule that applies to TV, and its timeline: the Commission "adopted rules on December 13, 2011" and they "became effective on December 13, 2012, one year after the date of their adoption." The core requirement reads:

"Federal Communications Commission (FCC or Commission) rules require commercials to have the same average volume as the programs they accompany."

The same page names the technical standard that broadcasters must follow, which is where your audio mixer gets the target:

"Specifically, the CALM Act directs the Commission to establish rules that require TV stations, cable operators, satellite TV providers or other multichannel video program distributors (MVPDs) to apply the Advanced Television Systems Committee's (ATSC) A/85 Recommended Practice ("ATSC A/85 RP") to commercial advertisements they transmit to viewers."

The FCC is also clear about the limits of the rule, and this passage is worth sharing with anyone who expects the law to make every ad equally loud:

"The ATSC practice that Congress directed us to adopt does not set an absolute cap on loudness. Rather, it requires commercials to have the same average volume as the programming they accompany, so that the volume a consumer chooses is the one at which both the programming and the advertisements will air."

It also notes the scope, answering "No, the CALM Act only applies to commercials aired on television." to the question of radio and internet ads. The Commission adds that "The rules should eliminate any systematic difference between the loudness of commercials and the loudness of the programming they accompany." It closes that answer with a cautious note: "We hope and expect that compliance with this practice will significantly reduce the problem of loud commercials for consumers."

ATSC updated A/85 in July 2026. It says the practice "provides guidance to broadcasters, streaming-media service providers, and creators of audio for television content" and that it:

"recommends methods to effectively maintain consistent loudness across content transitions, discusses audio metadata systems and their use, and describes modern audio dynamic-range control."

ATSC also publishes a short companion, the "Loudness and True Peak Quick Reference", described this way: "This document provides a summary of essential loudness values and measurement techniques from A/85, for quick reference and convenience." It is the page to hand your mixer.

In Europe and many other markets, the reference is EBU R 128, which the EBU summarizes in its publication notes as follows:

"Recommendation to use an average programme loudness of -23 LUFS and the 'Loudness Range' and 'Maximum True Peak Level' descriptors."

For viewers, the FCC route is a complaint, and its advice is to "Provide the FCC with specific information about TV commercials that seem louder than the programming they accompany by filing a complaint." For advertisers, the safer route is to avoid triggering one. Online platforms apply their own normalization, so the safest approach is to mix to the broadcast spec your media plan requires and ask the studio for platform versions as needed.

The delivery package to ask for

Ask for stems and the clean voice track even if you do not need them today. A new cutdown, a caption file or a translated version all start from those files, and recovering them months later is slow. For a fuller view of the finishing stage, our video post production guide covers editing, sound and color in sequence.

Common Mistakes Brands Make With Commercial Voice Over

• Casting before the script is locked. The voice you picked for a gentle 60 second story may be wrong for the punchy 15 that legal approves a week later.
• Booking for web and running on TV. Usage creep is the most expensive voice mistake, because it surfaces after the media has already run.
• Too many words. Overwritten scripts force rushed reads, and rushed reads sound like advertising in the worst sense.
• Directing by committee. Pick one director on the call and route all other notes through them.
• Ignoring pronunciation. A mispronounced product name in a national spot is a costly re-record.
• Skipping loudness checks. A mix that is not checked against the target spec risks rejection at the station or an unbalanced sound on streaming.
• Treating AI voices as rights-free. A synthetic voice still needs a license or consent, and contracts should say so.

How a Studio Like Neverframe Handles Voice in Ads

At Neverframe, voice over is built into the production from the script onward. Brands and agencies commission one of four formats, and voice is built into the ones that need it.

Spots come with a 4K master plus a vertical version, so the voice track is mixed for both the broadcast cut and the mobile cut. Films come with a 4K master and social cuts. The first version arrives in a few days, 72 hours for agencies, and final delivery is in two weeks, three to four for films.

In practice, that means the script is timed out loud before casting, the read is directed against an animatic rather than a blank page, and the usage terms for voice and music are aligned with your media plan before anything is recorded. If you are planning a spot with narration, tell us about the campaign and we will scope the format, the voice and the timeline with you.

Agencies can work with us white label, with partner pricing of up to 30% off. Details are on the partners page.

FAQ

What is a commercial voice over?

A commercial voice over is narration recorded for paid advertising, such as TV and streaming spots, pre-roll, paid social, radio and podcast ads. The voice speaks from off screen to carry the message, the offer or the legal copy. It differs from non-broadcast narration, like training or explainer videos, mainly in how it is priced: commercial work is priced on usage, meaning where the ad runs, for how long and on which platforms, because paid media gives the voice much wider exposure.

How many words fit in a 30 second commercial voice over?

A comfortable speaking pace sits around 150 words per minute, which gives about 75 words for a full 30 seconds. Once you allow for a musical open, a logo at the close and pauses between ideas, most 30 second spots work best with 55 to 65 words. A 15 second spot holds roughly half of that and usually needs its own script rather than a faster read of the 30. Always time the script out loud at the intended pace before booking talent.

How are commercial voice over rates calculated?

Rates for broadcast work are based on usage. The main factors are the media (TV, streaming, online, radio), the territory (local, national, international), the term (weeks, a year, a buyout), exclusivity and the number of spots and cutdowns. The talent's experience and union status also matter. Non-broadcast work is usually priced per word or per finished minute. Ask for renewal pricing in the first contract and list every planned cut at booking, since adding cuts later changes the usage.

Can brands use AI voices in commercials?

Yes, with consent and a license. Use a licensed synthetic voice or a clone created with the written agreement of the person whose voice it is, with terms that define media, territory and duration. Avoid any request for a voice that imitates a specific celebrity or actor. Tennessee's ELVIS Act added voice to its likeness protections, and the U.S. Copyright Office has recommended a federal law against unauthorized digital replicas. Put AI terms in human voice contracts as well, so training and synthetic use are agreed in advance.

What files should I receive after a voice over session?

Ask for a final mix to the broadcast loudness spec of your market, an online mix, separate stems for voice, music and effects, a clean edited voice track, labeled alternate reads, wild lines such as the brand name and call to action, session notes and a written usage summary. Stems and the clean voice track matter most, because every future cutdown, caption file or translated version starts from them. Recovering these files months after the session is slow and sometimes impossible.

Ready to Plan a Spot With a Voice That Fits Your Brand

The quality of a commercial voice over is mostly decided before the session: a script timed out loud, a casting brief that describes the listener, usage terms that match the media plan, and a mix built for the platforms where the ad will run. Neverframe produces films, spots, social ads and avatars for brands and agencies, with voice and sound built into the work. See the formats on our brands page or start a conversation about your next campaign.