Running your practice
Localisation Strategy: How Agencies Plan, Price, and Deliver Multi-Locale Programs
A localisation strategy defines which locales to launch, what lanes to translate, which quality tier each lane earns, and how the agency bills — scoped for
Every localisation program needs a planning layer that sits above the individual translation projects — the document that tells a producer which locales, which content lanes, which quality tier, and which workflow apply before a single resource bundle is opened. This page writes that layer from the agency's side of the desk, not the buyer's.
What we offer
What is a localisation strategy from the agency's side of the desk?
A localisation strategy is a written plan the agency and client sign before any string is translated, defining which locales to launch, which content lanes to localise, which quality tier each lane earns, how terminology and style are governed, and how the work is priced and measured. It is the agency's scoping artifact, not a marketing deck.
What are the core elements of a localisation strategy an agency must document?
A complete localisation strategy documents seven elements: target locales and their tier, in-scope content lanes, quality tier per lane, terminology and style governance, workflow and tooling, pricing basis per lane, and measurable KPIs. Missing any one of the seven turns the document into a wishlist rather than a signable SOW schedule.
| Element | Owner | Output in the signed document |
|---|---|---|
| Locale tiers | Agency strategist + client PM | Tier matrix (Tier 1–4) with named locales |
| Lane scope | Agency strategist | Lane × tier scope sheet |
| Quality tier | Agency locale lead | Error-per-1000-words threshold per lane |
| Governance | Agency + client reviewer | Term base / style guide ownership clauses |
| Workflow | Agency engineering | TMS integration diagram, pseudo-loc checkpoint |
| Pricing basis | Agency commercial | Per-lane rate card with TM leverage schedule |
| KPIs | Both parties | Four measured dimensions tied to monthly invoice |
Which target locales make the first cut, and how are they tiered?
Locales are grouped into four tiers: Tier 1 full localisation across all six lanes, Tier 2 product and support only, Tier 3 machine translation with light post-editing, Tier 4 English retained with locale-aware formats. The tier a locale earns drives every downstream cost.
| Tier | Lane coverage | Quality tier | Typical use |
|---|---|---|---|
| Tier 1 — Full | All six lanes | Publication-grade on legal and marketing; functional on product | Top-revenue or regulated markets (e.g. Japan, Germany) |
| Tier 2 — Partial | Product strings + support KB | Functional | Secondary growth markets |
| Tier 3 — MT + PEMT | Product strings only | Gist with post-editing | Long-tail markets validating demand |
| Tier 4 — English retained | None localised; locale-safe formats only | N/A | Markets where English is the de facto business language |
Which content lanes belong in scope, and which stay in English?
The six lanes — product strings, website, marketing, support KB, legal/compliance, and in-product video — are each scoped in or out per locale tier. Legal and compliance stays in-scope for regulated markets regardless of tier. The website lane is scoped per our Website Localization: What Translators and Agencies Actually Do on the Project playbook rather than re-explained here.
Scope assignment per lane follows three rules:
- Legal and compliance is always in-scope where the locale has regulatory exposure.
- Product strings are in-scope wherever a user can select the locale in-app.
- Marketing and video are in-scope only when the client funds in-market demand generation.
What quality tier does each lane earn, and who reviews it?
Quality tiers run from publication-grade through functional to gist, each with a named reviewer pool and a per-1000-word scoring rubric at release.
- Publication-grade — legal filings, marketing hero copy. Reviewer: in-country subject-matter expert. Threshold: under 2 errors per 1000 words.
- Functional — product strings, support KB top articles. Reviewer: senior in-country linguist. Threshold: under 5 errors per 1000 words.
- Gist — long-tail KB, user-generated content. Reviewer: post-editor spot-check. Threshold: comprehensibility pass on sampled segments.
Who owns the term base and style guide, and how are changes approved?
Term base ownership belongs to the agency's locale lead. Style guide ownership belongs to the client's in-market reviewer. Changes to either require written sign-off from both before the next sprint pulls the update.
The governance clause set the agency inserts into every strategy document covers five items:
- Term base ownership and change-request workflow.
- Style guide ownership and sign-off path.
- New-locale approval authority and lead time.
- TM leverage discount schedule.
- Quality-score release gate and its override procedure.
What are the measurable benefits a localisation strategy unlocks for the client?
Documenting the expected gains gives the strategy a business case that survives budget scrutiny. The four benefits the strategy underwrites are:
- Faster time-to-market per locale — one shared backbone cuts repeat scoping work on each new launch.
- Reduced rework costs — pseudo-loc and TM leverage surface defects before translator hours are spent.
- Higher in-country conversion rates — publication-grade marketing on Tier 1 locales outperforms raw translation.
- Defensible compliance posture — legal lanes carry a quality tier and reviewer sign-off that holds up in audit.
How does an agency build a localisation strategy step by step?
An agency builds the strategy in six steps: audit the source, tier the locales, scope the lanes, set the quality rubric, wire the workflow, and lock the pricing basis. The output is a signed document delivered before any resource bundle is translated.
Step 1 — How do you audit the source for localisation readiness?
Audit the source by checking for externalised strings, locale-safe date and currency handling, Unicode support, and a stable resource bundle format. Anything failing the audit becomes an i18n ticket owned by the client's dev team, not a translation line item. See i18n: What Internationalization Means for Translators Producing Localized Builds for the engineering prerequisites.
The audit output is a one-page defect list with four columns: defect, severity, owner, and target fix date. The strategy document cannot be signed until every high-severity defect has a named owner.
Step 2 — How do you tier the target locales?
Tier each locale by three inputs: revenue potential from the client's market research, regulatory requirements, and the agency's existing translator coverage and TM leverage for that language pair. Regulatory exposure forces Tier 1 regardless of revenue.
The tiering output goes into the strategy as a locked matrix. Changing a locale's tier mid-program requires a formal change order.
Step 3 — How do you scope each content lane per locale?
Scope lanes by multiplying the locale tier against the lane matrix:
- Tier 1 locales take all six lanes.
- Tier 2 takes product strings plus support KB.
- Tier 3 takes MT-plus-PEMT on product strings only.
- Tier 4 keeps English with locale-safe date, number, and currency formats.
Step 4 — How do you set the quality rubric and reviewer pool?
Set the quality rubric by assigning each lane a target error-per-1000-words threshold, a named reviewer pool, and a release gate that blocks publishing if the score falls below the threshold on sample segments. The rubric is carried over from our certified-translation QA model and adapted lane by lane.
Reviewer pools are named individuals, not vendor categories. The strategy lists the primary reviewer and one named backup per locale per lane.
Step 5 — How do you wire the workflow and tooling?
Wire the workflow by connecting the client's repo or CMS to the agency's TMS, defining pseudo-loc and staging checkpoints, and setting TM leverage rules so pre-translated segments route to review rather than full translation. Framework-specific pipelines (React i18n, Shopify, Webflow, Pendo) inherit from this one wiring diagram.
Where does machine translation fit as a production stage, not a finished answer?
Machine translation sits as a pre-translation stage for Tier 3 locales and long-tail KB content. It is always followed by post-editing against the same term base and style guide used elsewhere. Raw MT output is never shipped.
Pseudo-localization sits earlier in the workflow as a checkpoint that surfaces i18n defects before translator time is spent. It is a build-step, not a translation step.
Step 6 — How do you lock pricing and measurement?
Lock pricing by assigning each lane its own basis and tying monthly invoices to the quality score at release. One rate card does not fit six lanes.
| Lane | Pricing basis | TM leverage | Reviewer pool |
|---|---|---|---|
| Product app strings | Weighted word | Full schedule applies | Senior in-country linguist |
| Website | Per-source-word + engineering hours | Full schedule applies | In-country marketing reviewer |
| Marketing content | Per-source-word, transcreation surcharge on hero copy | Partial (new assets rarely leverage) | In-country marketing reviewer |
| Support KB | Weighted word, MT+PEMT on long-tail | Full schedule applies | Post-editor spot-check |
| Legal / compliance | Per-source-word + per-1000-words quality surcharge | None on new regulated text | In-country subject-matter expert |
| In-product video | Per-minute for subtitles; per-minute higher for dubbing | None | In-country voice/subtitle reviewer |
Pricing
How do you price a localisation strategy engagement itself?
The strategy document is priced as a fixed-fee pre-production engagement — scoped as a 2–4 week consulting block covering source audit, locale tiering, lane scoping, and the signed pricing matrix. It is delivered before any translation PO is cut. Giving the strategy away as pre-sales trains clients to treat planning as free.
Fixed-fee pricing avoids the two failure modes of hourly consulting: scope creep on the agency side and budget anxiety on the client side. The quote covers one audit cycle, one tier matrix, one scope sheet, one rubric, and one 12-month pricing schedule — revisions beyond that are change orders.
What goes into the fixed-fee pre-production deliverable?
The deliverable bundles six artifacts into one signed document:
- Locale tier matrix.
- Lane scope sheet.
- Quality rubric with per-lane thresholds.
- Governance clauses.
- Workflow diagram with TMS integration points.
- 12-month pricing schedule.
The signed document becomes the SOW schedule referenced by every subsequent translation PO. One strategy, many POs.
Where does regulated content (e.g. USCIS filings, EU legal) force a quality tier up?
Regulated content forces publication-grade quality regardless of locale tier. For example, USCIS-ready certified translation runs $0.10–$0.25 per source word or $25–$150 per page, with a certificate of accuracy signed by a competent translator required on every page. The strategy document must budget regulated lanes separately from standard localization lanes, because the TM leverage schedule does not apply to new regulated text.
The same quality-tier uplift applies to EU medical device labeling, financial disclosures, and privacy policies in jurisdictions with named-regulator exposure. The strategy names the regulator the lane is prepared for; the receiving authority confirms acceptance.
Get a quoteWhat KPIs does the strategy hold the translator and agency to?
The strategy holds the agency to four measurable KPIs per release, each tied to monthly invoicing:
| KPI | Measurement | Target | Invoice impact |
|---|---|---|---|
| Quality score per 1000 words | Weighted rubric on sampled segments | Below lane threshold | Release gate; miss blocks publish |
| TM leverage % | Leveraged words / total words | Baseline set in Step 6 | Discounts applied against full rate |
| On-time delivery | Delivered vs sprint cadence | 95% on-time | Service credits below target |
| In-locale reviewer sign-off rate | Approved releases / submitted releases | 90% first-pass approval | Repeat rework at agency cost |
What are the common failure modes in a localisation strategy, and how do agencies prevent them?
Five failure modes recur, each prevented by a specific governance clause in the strategy document:
- Launching Tier 1 locales without regulated-content budget. Prevented by a legal-lane line item on every Tier 1 locale in the pricing schedule.
- Treating the term base as a one-time deliverable. Prevented by a term base ownership clause assigning the agency locale lead as permanent owner.
- Skipping pseudo-loc. Prevented by a workflow clause making pseudo-loc a mandatory sprint-zero checkpoint.
- Pricing all lanes on raw word count. Prevented by the per-lane pricing basis table in Step 6.
- Letting the client's PM own the style guide. Prevented by a style guide ownership clause naming the in-market reviewer, not the PM, as owner.
How does a localisation strategy connect to adjacent delivery lanes the agency already runs?
The strategy document references the agency's existing lane playbooks rather than re-explaining them. Each lane inherits a single consistent process from the signed strategy and the published guide.
- Application lane — follows the Application Localization: The Translator's and Agency's Guide.
- Website lane — follows the Website Localization: What Translators and Agencies Actually Do on the Project playbook.
- Engineering readiness — follows i18n: What Internationalization Means for Translators Producing Localized Builds.
Framework and platform specifics — React i18n pipelines, Shopify and Webflow CMS routing, Pendo in-product content, and locale-specific guides like Chinese localization — carry their own siblings that the strategy document names rather than duplicates.
The questions agencies hear most on strategy engagements cluster around examples, when a localisation strategy fits versus a transnational one, and how it differs from international strategy and from a translation plan.
Frequently asked questions
What is a localization strategy?
A localization strategy is a structured plan for adapting content, products, or services to meet the linguistic, cultural, and regulatory requirements of specific target markets. For translation agencies and LSPs, it defines which locales to support, how workflows are staffed and tooled, how certified translation programs are priced, and what quality standards apply per locale. A mature strategy accounts for volume forecasting, translator specialisation (legal, medical, financial), technology stack, and turnaround SLAs — typically broken down market by market rather than applied globally as a single ruleset.
Can you give me an example of localization?
A practical example for a translation agency is delivering certified legal translations for a client expanding from the US into the EU: the same source contract is adapted into German, French, Polish, and Spanish, each version carrying a sworn or certified translator's stamp valid in that jurisdiction. Beyond language, the process may require adjusting date formats, legal terminology, notarisation conventions, and even document layout to meet country-specific official requirements. Each locale is treated as a distinct deliverable, not a copy-paste variant.
When should a firm follow a localization strategy?
A firm should adopt a formal localization strategy when it is serving — or planning to serve — multiple language markets where compliance, cultural fit, or legal certification requirements differ meaningfully. For translation agencies, this threshold is typically crossed when managing more than two or three active locales simultaneously, when certified output (USCIS, sworn, apostilled) is part of the deliverable mix, or when clients demand consistent turnaround and pricing across markets. Ad-hoc locale management becomes a quality and margin risk beyond that scale.
What is global localization?
Global localization (sometimes called 'glocalization') is the practice of designing a program that scales across many markets while still customising output for each individual locale's linguistic and cultural context. For LSPs running multi-locale certified translation programs, it means standardising the production workflow — intake, first-draft generation, translator review, certification — while allowing locale-specific rules (target language conventions, required credentials, stamp formats) to govern the final deliverable. The goal is operational efficiency at scale without sacrificing jurisdiction-level compliance.
What is internationalization and localization?
Internationalization (i18n) is the process of designing a product or workflow so it can support multiple locales without requiring structural redesign; localization (L10n) is the act of adapting that output for a specific market. In a certified translation production context, internationalization means building a workflow — including first-draft tooling, style guides, and QA checklists — that can accommodate any language pair and certification format, while localization means applying the correct target-language terminology, legal register, and certification conventions for each jurisdiction delivered.
What is internationalization vs localization?
Internationalization is the upfront infrastructure work; localization is the market-specific execution that runs on top of it. A translation agency internationalizes when it configures its TMS, CAT tools, and first-draft pipeline to handle bidirectional scripts, varied certification templates, and multi-currency pricing. It localizes when a translator adapts a birth certificate for the German standesamt versus the Spanish registro civil — applying the exact terminology, format, and credential requirements each authority demands. Both disciplines are necessary; neither replaces the other.
What is L10n?
L10n is the standard industry abbreviation for 'localization,' formed by taking the first letter 'L,' the count of omitted letters (10), and the last letter 'n.' It is widely used in LSP project management, TMS interfaces, and tool documentation to save space. In certified translation production workflows, L10n tasks specifically include adapting linguistic content, legal register, certification wording, and document formatting to meet the requirements of a target jurisdiction — distinguishing this work from the broader i18n (internationalization) setup that enables it.
What are the best localization tools?
The most widely used localization tools in professional translation production are CAT (computer-assisted translation) platforms such as SDL Trados, memoQ, and Phrase, combined with translation memory and terminology management systems. For certified translation programs specifically, agencies also rely on TMS platforms (XTRF, Plunet) for project and vendor management, and increasingly on AI-assisted first-draft generation tools to reduce per-word production costs before a human translator applies certification review. Quality assurance tools like Xbench or Verifika are standard for pre-delivery checks across large multi-locale batches.
What are some tips for becoming a good translator?
The foundation of a good professional translator is near-native command of both source and target language combined with deep subject-matter expertise in at least one specialist domain — legal, medical, or financial. For certified translation work specifically, understanding the certification and notarisation requirements of each target jurisdiction is as important as linguistic accuracy. Productive translators also invest in CAT tool proficiency, maintain curated termbases, and treat post-editing AI first drafts as a distinct skill: knowing when to accept, revise, or override machine output without losing speed gains.
Related
- Localization Codes: The Standards Translators and Agencies Use
- Localization Coding: How Translators and Developers Share the Resource File
- Localization Engineer Job: What the Role Does, Pays, and Requires
- Localization Files: What Translators Receive, Parse, and Return
- Localization Management: How Translators and Agencies Run Multi-Locale Programs
- Localization QA: How Translators and Agencies Run LQA Before Release