Translation vs localization

Translation converts words. Localization adapts content to a new culture. For international SEO, localization is the work that separates sites that rank in a market from sites that are just accessible there.

Translation, the surface layer

Takes source text, renders it in the target language. Can be done by:

• Machine translation (Google Translate, DeepL)
• Translation software (SDL, memoQ)
• Human translators

Translation preserves meaning. It doesn't adapt to the target market.

Localization, the deeper layer

Localization adjusts:

• Language + dialect. UK English vs US English, Castilian Spanish vs Mexican Spanish
• Cultural references, examples, analogies, humor that land locally
• Currency + pricing, local currency, appropriate price points
• Date + number formats, dd/mm/yyyy vs mm/dd/yyyy, decimal vs comma separators
• Units, metric vs imperial
• Addresses + phone formats, country-specific conventions
• Images + imagery, culturally appropriate photos, relevant examples
• Legal + regulatory. GDPR notices, local consumer protection language
• Trust signals, local awards, local press mentions, local partnerships
• Keyword targeting, what people actually search in that market (often different from direct translations)

Why machine translation fails for SEO

1. Keyword mismatch

The direct translation of your English keyword might not be what locals search. Example: "running shoes" → "chaussures de course" in French, but many French speakers search "baskets" or "sneakers."

2. Awkward phrasing

Even good MT produces sentences that read unnaturally. Google's NLP (and users) detect it.

3. Missing context

MT doesn't know your industry, audience, or brand. Produces generic, flat translations.

4. Duplicate content risk

Multiple sites using MT on the same source content produce similar translations. Google may treat them as near-duplicates.

5. E-E-A-T hit

Unnatural text reads as low-quality. Google's Helpful Content signals punish it.

The localization workflow

Step 1: Native speaker translation

Start with human translators who are native speakers of the target language. Ideally with subject matter expertise.

Step 2: Local keyword research

Don't assume translated keywords are the ones people search. Do local keyword research per market. Adjust titles, H1s, content to target real local searches.

Step 3: Cultural adaptation

Examples, case studies, screenshots, analogies, adapt to local context. A US case study about Thanksgiving means little in Germany.

Step 4: Practical adjustments

Currency, dates, units, phone formats, addresses. Every touchpoint that feels "foreign" to local users gets adjusted.

Step 5: Legal review

GDPR for EU, various consumer protection laws globally. Privacy policies, terms, cookie banners all need local review.

Step 6: Local trust signals

Local awards, partnerships, press, testimonials, certifications. Add to the market-specific pages.

Step 7: Local SEO basics

If you have a local presence in-market: local Google Business Profile, local directories, local backlinks.

Hybrid approach that works

1. Use MT for first draft (fast + cheap)
2. Native-speaker editor rewrites for natural flow
3. Local marketer adds keyword targeting + cultural adaptation
4. Legal review for regulatory fit

Result: quality close to full human translation, at a fraction of the cost.

Tools that help

• DeepL, best MT for many European languages
• Phrase / Lokalise / Crowdin, translation management platforms
• Weglot, easier plugin approach for small sites (trade-offs apply)
• Local SEO tools. Ahrefs/SEMrush filtered to target market for keyword research

Common localization mistakes

• Treating all Spanish markets as one (Spain ≠ Mexico ≠ Argentina)
• Using Google Translate as final output
• Forgetting to localize currency + pricing
• Translating blog content but not product/category pages
• Keeping US images in EU versions of the site
• Translating but not doing local keyword research
• Using hreflang to point to unlocalized pages ("we translated this once, we're done")