If a product video is going onto an e-commerce site serving EU consumers, the procurement question is not simply whether to burn subtitles into the final cut. The real question is whether a customer can obtain the same purchase-critical information when they cannot hear the soundtrack, cannot see key visual cues, or cannot operate the player with a mouse.
That does not mean every brand video needs the same accessibility package. It means the brand should classify the content, the channel and the information risk before deciding whether it needs captions, a transcript, visual description, a separate described version, an accessible player or an equivalent text presentation on the page.
Separate the legal scope from the production method
The European Commission’s overview of the European Accessibility Act lists e-commerce among the covered services. Directive (EU) 2019/882 applies to relevant products placed on the market and consumer services provided after 28 June 2025.
The Directive’s general service requirements include providing information through more than one sensory channel, making information available in text formats that can generate alternative formats, supplementing non-text content with an alternative presentation, and making websites and mobile services perceivable, operable, understandable and robust. It also contains exclusions, including certain prerecorded time-based media published before 28 June 2025, uncontrolled third-party content and archives that are no longer updated. The EU’s business guidance also describes exceptions for microenterprises providing services and the possibility of a documented disproportionate-burden assessment.
These facts make product video part of a wider accessible e-commerce experience. They do not create a universal instruction that every advertisement must carry one identical subtitle file. National implementation, the operator’s status, the service scope, channel capabilities and possible exemptions all matter. Brands should confirm legal applicability with qualified advisers in the markets they serve.
The W3C Web Content Accessibility Guidelines provide more specific technical benchmarks. WCAG 2.2 calls for captions for prerecorded audio in synchronized media at Level A. It also addresses a media alternative or audio description for visual information at Level A, and audio description for prerecorded video at Level AA. WCAG is a valuable implementation and acceptance reference; it should not be misrepresented as a single EAA file checklist for every commercial video.
Translation subtitles and accessible captions are not the same deliverable
Cross-border teams often collapse several different assets into one line item called “subtitles.”
| Deliverable | What it solves | What it can still miss |
|---|---|---|
| Translation subtitles | Makes dialogue understandable to people who can hear but do not know the source language | Speaker identity, music, alarms and meaningful sound cues |
| Accessible captions | Presents dialogue and meaningful audio information in synchronized text | Important visual-only information |
| Basic transcript | Creates searchable, assistive-technology-readable text for the soundtrack | No timing and often no description of the picture |
| Descriptive transcript | Records important sound and visual information | Does not replace an accessible playback experience in every channel |
| Audio description | Speaks important visual information that the main soundtrack does not carry | Needs script space, mixing and player support |
| Accessible player | Lets people operate playback and select access tracks | Does not create accurate captions or descriptions by itself |
Imagine a headphone setup video where a red status light turns white to show that pairing is complete, while the voice-over only says, “You are ready to go.” Perfect captions would still omit the pairing state for someone who cannot see the light. If the script instead says, “When the light changes from red to white, pairing is complete,” the critical visual fact is already present in the main audio and the additional description burden may be reduced.
This is why accessibility belongs in the script and storyboard. Post-production can generate words, but it cannot easily create pauses inside a locked 15-second commercial without changing its rhythm or duration.
Let purchase information determine the delivery tier
The useful question is not, “Should every video receive the maximum package?” Ask instead: “If the sound or picture disappears, what purchase information does the customer lose?”
Light tier: the moving image is mainly decorative
For a silent decorative loop that does not independently communicate specifications, operation or safety information, the product page may already carry the substantive information in text. The priorities are avoiding unwanted audio, ensuring the player does not block operation, and confirming that the video is not the only path to understanding the offer.
Information tier: the soundtrack explains the product
When dialogue, narration or an alert sound is necessary to understand a feature, provide synchronized captions that have been reviewed by a person. Accessible captions identify speakers and meaningful sounds; they are not merely translated dialogue. A clean transcript can also support the product page, customer service, search and later localization.
Decision tier: the picture carries the buying judgment
Installation order, component position, size differences, indicator lights, interface changes and usage restrictions may exist only in the image. Put those facts into integrated narration, a descriptive transcript or an audio-described version. The right choice depends on the video length, whether the channel supports alternate audio, and whether the page offers an equivalent text explanation.
High-consequence tier: safety, health, finance or regulated claims
If the image communicates a safety warning, eligibility condition or another high-consequence decision, add market-specific legal review, a terminology record, version freezing and documented approval. Automated captioning can start the workflow; it should not be the final authority for numbers, product names, warnings or synchronization.
ONCE ACCESS: write accessibility into procurement
ONCE ACCESS turns accessibility from a launch-day repair into six production decisions.
A — Assess purchase-critical information
Mark information that exists only in sound, only in the picture, or in both. Prioritize specifications, installation, compatibility, status, limits, warnings and next actions. Do not waste the description budget narrating every decorative camera move.
C — Choose a sensory alternative
Use captions for speech and meaningful sound. Use integrated narration, a descriptive transcript or audio description for essential visual action. Pair those assets with accessible player controls and page text where the channel permits.
C — Create space at script stage
When important visuals can be named naturally, write them into the primary script. If a separate described track is needed, reserve pauses between dialogue and music. Define caption-safe areas before layout and aspect-ratio adaptation so captions do not cover the product, price or action control.
E — Export for the channel, not just as an MP4
A practical package can include a clean master, human-reviewed SRT or WebVTT captions, a clean transcript, a descriptive transcript, a described version when needed, and a manifest showing the intended market, language and channel for each asset. Social platforms may require open captions; a brand website will often benefit from user-controlled closed captions. One export should not be copied mechanically everywhere.
S — Sign off through real use
Run at least three checks: watch without sound to find missing audio information; listen without the picture to find missing visual information; and use keyboard or assistive technology to verify the player and transcript. After automatic transcription, check names, numbers, product terms, meaningful sound and timing.
S — Sustain every version
Price, specification, narration, edit or market-language changes can invalidate captions and descriptions. Tie the master, captions, transcript and described assets together in the version record so an updated video is not served with an outdated accessibility layer.
Measure the full cost per deployable combination
Per-minute caption pricing captures only one visible task. A more useful model is:
Total accessible delivery cost = base production + accessibility design + captions and transcripts + description or alternate versions + language review + channel QA + version maintenance
Divide that total by the number of approved market × language × channel × version combinations. This produces a full cost per deployable combination and allows a brand to compare a cheap master that repeatedly requires remediation with a better-planned asset system that can move reliably across markets.
Change the brief before buying more caption software
Light treatment may be enough when a video is decorative and the page already provides equivalent information. Translation subtitles alone are usually inadequate when the video carries product explanation, installation, compatibility or safety decisions.
For the next request for quotation, give the production partner the release countries, channels, player capabilities, purchase-critical information and intended accessibility target. Then price the correct access layer.
The durable deliverable is not “a subtitled final cut.” It is a product-content package that knows what must be heard, seen, read, operated and kept current.
Sources and scope
- European Commission: European Accessibility Act, accessed 11 October 2026.
- Directive (EU) 2019/882, including the scope for e-commerce services and Annex I accessibility requirements.
- Your Europe: Services, goods and their digital accessibility, accessed 11 October 2026.
- W3C WAI: Making Audio and Video Media Accessible, including WCAG 2.2 Success Criteria 1.2.2, 1.2.3 and 1.2.5.
- Greycrest Visuals: Accessible Video Production: A WCAG Guide for Organizations, July 2026. This peer guide was used to compare procurement practice; legal scope is based on the primary sources above.
This article offers production and procurement guidance, not legal advice. Confirm obligations against the law of the relevant Member State, the business’s status, the publishing channel and qualified counsel.
