Planning What Every Language Gets Across Web, Docs, Email, Social and Video

What the strategy has to decide

Most organizations start translating content one request at a time: a sales team asks for a German brochure, support asks for Spanish help articles, someone subtitles a launch video in Japanese. Each decision makes sense alone. Together they produce a site where a German visitor reads a translated homepage, clicks through to English pricing, receives an English onboarding email and lands on Spanish-only help content by mistake.

A cross-channel strategy replaces those one-off decisions with a few written answers:

• Coverage: which content types each language receives, and at what depth.
• Sequence: what gets translated first, so every translated page leads somewhere that is also translated.
• Consistency: one set of terms, one voice and one source of truth shared by every channel.
• Resourcing: who translates, who reviews and who updates each content type in each language.
• Maintenance: what happens when the source changes, and when a translation is retired instead of updated.

Video is one row in this plan, not the whole plan. The detailed decisions for a video library (auditing it, tiering it, choosing subtitles or voice) belong in a video localization strategy; this article is about how video sits alongside everything else.

Inventory content by type, not by channel

Start with a list of content types, because types behave differently from channels. A product page and a pricing page both live on the website but age at very different speeds. A list that works for most teams:

Website core pages

Homepage, product and pricing pages; low volume, high visibility, updated with each launch

Long-form marketing

Blog posts, guides, case studies; high volume, long shelf life, varied value

Documentation and help

Help articles, API docs, release notes; high volume, changes with every release

Lifecycle email

Welcome, onboarding, billing and renewal emails; few templates, sent to every customer

Social and campaigns

Posts, ads, short videos; short shelf life, often better written locally

Video

Explainers, demos, tutorials, webinars; expensive to change once published

Legal and transactional

Terms, privacy notices, invoices, receipts; usually requires qualified translation

For each type, note four things: volume, how often it changes, how visible it is, and whether a mistake carries legal or financial risk. Those four attributes drive almost every later decision.

Decide what each language gets

Not every language needs every content type. Define three or four coverage levels and assign one per language:

1. Full: every content type, kept current. Reserved for markets that bring significant revenue or have legal requirements.
2. Core plus support: website core pages, lifecycle email, the most-used help articles and key videos. Long-form marketing stays in the source language.
3. Entry: website core pages and a translated landing page per campaign, with a clear note that support and documentation are in the source language.
4. Discovery only: translated metadata, video subtitles and perhaps a single landing page, used to test interest before committing.

Choosing the languages themselves is a separate exercise based on your traffic, customers and sales presence, covered in which languages to translate videos into. The strategy only needs to record which coverage level each chosen language sits at and why.

A coverage matrix makes this visible. Put content types in rows and languages in columns, then write "full", "top 20 articles", "subtitles only" or "source language" in each cell. When a request arrives, the matrix answers it: if Japanese is at the entry level, a request to translate the whole blog is a strategy change, not a routine task.

Sequence so nobody hits a dead end

Translate in the order a reader moves through your content, so each translated piece links to another translated piece. A common mistake is translating a campaign ad and landing page while the sign-up flow and welcome email remain in English; the visitor converts and is immediately dropped back into the source language.

A sequencing rule that holds up in practice:

1. Core website pages and the navigation that connects them.
2. Anything a new customer is sent automatically: sign-up screens, welcome and billing emails.
3. The help articles and videos behind the most common questions from that market.
4. Campaign content, once there is a translated destination to send people to.
5. Long-form marketing and archive content, selectively.

Where a step cannot be translated yet, say so on the page. "Our help center is currently in English" is better than a link that silently switches language. The full customer-journey view of this ordering is covered in a separate article; here the point is simply that sequence is a strategy decision, not something to leave to whoever asks first.

Keep every channel saying the same thing

Consistency problems multiply with channels. The help center calls a feature "workspaces", the dubbed demo says "projects", and the email says "teams" because three people translated three pieces with no shared list.

Three shared assets prevent most of this:

• A glossary with approved translations for product names, feature names, plan names and terms that stay in the source language. Every translator, reviewer and tool should use it.
• A style guide per language covering formality, punctuation, number and date formats, and how to address the reader. The article on creating a translation style guide walks through what to include.
• A source-of-truth record listing each translated asset, the source version it was made from, and its owner.

Video creates a particular consistency risk because spoken terms are hard to search. Keeping a transcript and subtitle file for every translated video lets you search for an outdated term across the video library the same way you would across help articles.

Resource each content type realistically

Translation cost is rarely the largest line. Review, publishing and keeping content current usually cost more, and they scale with the number of languages times the rate of change.

Estimate per content type and language:

• Initial translation: machine draft, professional translator, or local writer.
• Review: who reads it, and how long that takes per page or per minute of video.
• Publishing: who uploads, formats and links it in each system (CMS, help desk, email tool, video host).
• Maintenance: how many updates per year the source receives, multiplied by the time per update.

Documentation is the usual surprise. It changes with every release, so a fully translated help center in four languages can need more ongoing hours than every other content type combined. That is why many teams translate only the most-viewed help articles and keep release notes in the source language.

Where video fits in the wider plan

Video sits awkwardly in content plans because it is expensive to re-record but cheap to translate in its current form. That makes it a good candidate for the "discovery only" level: subtitles or a translated transcript let you test whether a market engages before you commit to a dubbed version or a locally produced one.

Some practical rules for the video row of the matrix:

• Match the video's language level to the page it sits on. A dubbed German demo on an English-only product page will rarely be found.
• Publish a transcript on the page in the same language, so the content is readable and searchable.
• Decide per video type whether subtitles or a translated voice serves the viewer; the subtitles vs dubbing guide sets out the trade-offs.
• Record which source version each translated video came from, because video is the content type most likely to drift unnoticed.

Common mistakes and trade-offs

• Translating by request rather than by plan, which leaves every language with a random subset of content.
• Treating all languages equally. Giving every language full coverage spreads review capacity so thin that quality drops everywhere.
• Translating high-volume, fast-changing content (documentation, release notes) before low-volume, high-visibility content (core pages, emails).
• Building separate glossaries per channel, so the same feature has three names in one language.
• Counting translation cost but not maintenance. A translated page that is never updated becomes a liability within a year.
• Skipping a retirement rule. Some translations should be withdrawn when the source changes rather than updated, especially for low-traffic content.

The main trade-off is breadth against depth. A strategy that covers five languages at entry level reaches more people than one that covers a single language fully, but each visitor gets a thinner experience. Neither is right in general; it depends on whether your goal is testing demand or serving customers you already have.

Where mydubly fits in a cross-channel plan

mydubly handles one part of this plan: spoken content. It takes a recorded video or audio file from your device and produces transcripts, SRT and VTT subtitles, and optionally a dubbed version with a translated voice track, in 21 languages. The video localization page describes that workflow.

It does not translate website pages, documentation, emails or social posts, and it does not translate on-screen text inside a video. It also has no URL import, API or content-management integration, so it fits as a manual step in your video row rather than as a connector to your CMS.

For planning, the useful facts are cost and output. A transcript or subtitle job, optionally translated, costs 1 credit per minute with a 5-credit minimum, so a ten-minute webinar costs 10 credits (1¢). A dub costs 50 credits per minute with a two-minute minimum, so the same webinar dubbed into one language costs 500 credits (50¢). Each job produces one target language, and a dubbed job also gives transcripts in both languages, which helps your glossary owner check terms against the rest of your content.

A next step you can take this week

List your content types, add the languages you already publish in, and fill in the coverage matrix with what actually exists today. The gaps will be obvious: translated pages that lead to untranslated ones, or a language with fragments of everything and complete coverage of nothing. Pick one language, agree its coverage level, and fix the sequencing gaps first. For the calendar side of planning, a multilingual content calendar helps schedule the work once the matrix is agreed.