Why did the quote gain another line item when the voice actor changed just one product name? This is a common point of disagreement in voiceover pricing discussions. Brand video voiceover costs may appear to be calculated by word count or duration, but the first question is whether the usage has changed: a website introduction, organic social media, and paid advertising do not necessarily purchase the same set of services for the same voice recording.
This article is for brand teams preparing an overseas website, YouTube channel, conference presentation, or advertising campaign. We break the same English sentence into two purchase orders: one that only solves the problem of recording a usable voiceover, and another that also covers media usage, term, territory, pickups, and multilingual masters. After reading, you can judge whether two English voiceover quotes are truly comparable and include commonly missed items in your request for quotation and acceptance criteria.
Purchase Order A: The Brand Buys a Usable Voiceover First
The first order has a straightforward goal: obtain a clean human voice that matches the picture's rhythm, uses correct pronunciation, and has the right emotion. It usually includes voice selection, auditions, directed recording, basic editing, and delivery in the agreed format. It does not mean the recording can be used through every channel, in every territory, and for every term.
Provide three things before requesting a quote. First, a script locked down to the sentence level, with product names, models, abbreviations, numbers, and units marked. Second, a timecoded picture reference showing the actor where each line enters and where it ends. Third, an emotional reference using performable directions such as calm and explanatory or credible with a sense of action, while avoiding vague adjectives such as a little more premium or a little more international.
There is no need to ask the actor to read the entire film for free during the audition. Three lines are more effective: the shortest tests the onset, the longest tests breath control, and the most difficult tests product names and terminology. After selection, organize the product names into a pronunciation guide, including the brand-approved pronunciations and stress. This keeps casting separate from copy changes and avoids discovering in the booth that a model number was pronounced incorrectly.
If a supplier provides only one total price, the buyer can ask for four verifiable deliverables: the number of candidate voices, the number of formal recording sessions, the number of usable takes, and the file specifications. More candidates do not necessarily mean better results; the key is that each candidate fits the intended age, language accent, and brand character. Formal recording should not be accepted by how many words were read, but by whether it leaves director-approved performances that can be placed on the timeline. Clarifying these points puts two suppliers on the same pricing basis.
Purchase Order B: The Voice Also Enters a Defined Media Usage Plan
The second order does not just buy recording labor. It must also state where the voice will appear, which markets it will target, how long it will be used, and whether it will run in paid advertising. Organic publication on a brand website and sustained placement across multiple territories should not both be vaguely covered by the words online use.
SAG-AFTRA Commercials Contract Materials They distinguish traditional digital use, social platforms, YouTube, and different usage methods, and also list situations in which replacing dialogue or scenes may trigger additional recording fees. They do not automatically apply to production projects in China, but they give buyers a clear reference: recording sessions and media usage are two separate matters, and the final standard should be the parties' written agreement.
Voice actor usage rights should therefore be written in at least one readable line, for example: English voiceover for the brand website, the brand's owned social media, and designated advertising accounts, with territory, term, and permission to cut shorter versions listed separately. If a distributor may change, an overseas agency may take over, or additional advertising markets may be added, clarify in advance whether the rights can be extended instead of trying to guess after the film goes live.
The usage plan must also correspond to the edit versions. Turning a sixty-second website film into a fifteen-second advertisement is not simply a matter of reducing duration: the voiceover may be rearranged, the tone may change, or an entirely new opening line may result. List the expected long, short, horizontal, and vertical versions when requesting a quote. You do not need to buy every unconfirmed use upfront, but you should agree on how new versions will be priced. When the brand adds assets later, it can distinguish between reusing the recorded voice, re-editing lines, and bringing the actor back into the booth.
Both Purchase Orders Need the Same Directed Recording Session
A price difference does not mean the director can be removed from the session. What matters more is whether the hour leaves enough usable performance layers for editing. A common approach is to record one uninterrupted reference pass, then work through pacing and stress line by line. For key lines, keep three clear directions such as restrained, direct, and warm, without asking the actor to cycle endlessly through a dozen emotions. The director logs the selected take number for each line, so the editor does not have to listen to the entire session again.
During recording, document the microphone, preamp gain, the actor's mouth-to-mic distance, and seated posture, and capture a section of clean room tone. After the main session, first export a 48kHz, 24-bit dry WAV for post-production while retaining a reference mix for listening. If only one word needs to be replaced a week later, these records help the engineer recreate the voice distance and room tone. Otherwise, even when the old and new words come from the same actor, they may sound as if they were spoken in two different rooms.
When the picture already contains dialogue, ambience, or product sound cues, run both a voiceover-only listen and a full-mix listen. The isolated pass checks sibilance, plosives, and mispronunciations; the full mix determines whether music or machine sounds mask the voice. For how voice direction corresponds to the emotion of the finished film, see ONCE's KTC Emotional Advertising Case Study; the point is not to imitate a particular voice, but to see how voice, performance, and rhythm work together to express the brand.
Assign one clearly designated approver for the directed recording day. If the brand, agency, and production team all give line-by-line direction on the same call, the actor will receive conflicting requests. A more reliable approach is for the brand to confirm informational accuracy first, after which one voice director translates the feedback into pacing, intensity, phrasing, pauses, and emotional actions. Resolve one issue per round, and immediately mark takes as preferred, backup, or unusable after recording. This prevents the team from returning later to dozens of files to hold another vote.
Changing One Word: When Is It a Pickup and When Does It Require a New Session?
Voiceover pickups are a frequent source of disputes because changing just one word can have entirely different causes. If the actor misread it or the recording file was damaged, the production side should normally correct it within the agreed scope. If the brand changes a product name, specification, or promotional message after script lock, that is a script change. If a piece originally intended only as a website introduction later becomes a faster-paced paid advertisement, the performance direction and usage plan may have changed as well.
Do not write only includes one revision in the purchase order. Specify the trigger conditions: how many lines are included per revision, how many days after delivery requests may be made, whether the original voice and director's direction will be retained, and whether anything beyond the scope is charged by pickup session or by new version. Also explain what happens if the actor's availability cannot be replicated. This allows the production team to retain the original session files, microphone setup, take logs, and room tone for pickups, while the brand knows which changes will genuinely increase costs.
A practical sign-off step is for the brand to confirm the wording, pronunciation, and number readings together before recording. Number the voiceover lines in the picture script as VO01 and VO02. Pickup files should retain those numbers and add a version number instead of naming ten files final. This directly reduces wrong replacements, missed replacements, and mixed-up multilingual versions.
Before a pickup, place the old finished version and the original selected takes on the same timeline, then check the breaths, speaking rate, and musical cues before and after the new word. If the word count of the sentence changes substantially and a local replacement would leave an obvious seam, rerecord the entire sentence. If a product name change also changes the stress, the preceding and following lines may need to be rerecorded together. Make this decision before the actor enters the booth; otherwise, the editor may discover only after the actor has left that the word is correct but the sentence no longer connects.
Multilingual Delivery Is Not a Copy of the Unit Price; It Must Be Realigned to Picture
Chinese, English, and German differ in sentence length and word order, so the same time window in the visuals may not accommodate the same sentence in every language. A multilingual audio track budget should cover not only the actors, but also translation review, a terminology glossary, a native-language director, duration adaptation, subtitle checking, and final mixing. Recording one rhythm sample before batch recording is more reliable than forcing every language to speed up after all recordings are complete.
YouTube Official Multilingual Audio Track Guidance allows channels with the relevant feature to upload different language audio tracks for the same video and requires the additional audio files to be approximately the same length as the video. This means delivery cannot consist of a pile of isolated lines: each language should be delivered as a complete track starting at 00:00 and matching the master length, with the language, mix version, and date identified in the filename.
RWS Video and Audio Localization Services Overview also lists transcription, translation, subtitles, and voiceover as separate services. For brand procurement, the key point is not to mistake translated for ready to record: product-name pronunciation, locally familiar units, spoken duration, and picture actions still need to be confirmed line by line for each language version.
Accepting Both Purchase Orders: Listen Three Times, Not Just for Vocal Timbre
First, turn off the music and listen only to the voiceover. Compare every line against the locked script, pronunciation guide, and take log, checking for wrong words, swallowed sounds, abrupt edits, and changes in room tone. Second, turn on the full mix and listen through a phone speaker and your usual headphones to determine whether music, product sounds, or dialogue cover the key selling points. Third, place the finished film in its real usage context and check whether the website autoplay, YouTube multilingual audio track, or short advertising version has selected the correct file.
The delivery checklist should include the mixed master, dry voiceover, music-and-effects version, complete audio tracks for each language, subtitle files, the pronunciation guide, and the final script. Filenames can follow the order project_language_usage_mix status_version_date. For how live, synthetic, and cloned voices should first be assessed for identity and platform suitability, see AIGC Advertising Voiceover Selection Test. Do not force different questions into the same purchase order.
When comparing suppliers, first determine whether the two orders are buying the same thing, and only then compare the total price. The actor's profile, word count, or recording hours are each just one dimension. What determines whether the voice will remain useful later is whether the usage plan, director sessions, pickup conditions, language masters, and playback testing in the intended publishing context are clearly defined.
Send ONCE the shortest, longest, and most difficult lines from the script, together with the channels, territories, terms, and language versions. We will first organize a one-page voice procurement sheet based on the Brand Video Production Solution and then determine which costs should be combined and which must be listed separately.
