Why Doesn't a Direct Arabic Translation of Your English Site Rank?

The problem

Most "Arabic SEO" work we're brought in to fix is really English SEO with the words swapped: an English keyword list run through translation software, English page structure kept as-is, and an RTL stylesheet bolted on at the end. Arabic is a root-and-pattern language — a single three-letter root can produce dozens of related surface words (verb forms, nouns, plurals, adjectives) that look nothing alike to a keyword tool but represent the same search intent to a human reader. A literal translation captures one of those forms and misses the rest of the demand, while the content itself reads as translated rather than written for the reader, which both users and search engines can tell.

Symptoms

• The Arabic version of a page ranks for far fewer keywords than the English version, even though the content covers the same topic.

• Google Search Console shows Arabic queries with real impressions landing on the English URL instead of the Arabic one, or vice versa.

• Arabic content reads stiffly or uses phrasing a native speaker wouldn't choose — a translation "smell" that increases bounce rate even when the information is technically correct.

• Keyword research misses obvious search volume because only the Modern Standard Arabic (MSA) term was checked, not the dialectal or Arabizi variants people actually type.

• RTL pages have layout bugs — icons, arrows, or spacing that were mirrored for English's left-to-right assumptions instead of being built with logical (start/end) properties from the start.

How we diagnose it

We start Arabic keyword research from the root, not the surface word — mapping the root-and-pattern family around a core concept so we capture verb forms, noun forms, and common derived words a direct-translation keyword list would miss entirely. We then check how the target market actually searches: Modern Standard Arabic for formal/informational queries, local dialect for commercial and conversational ones, and Arabizi (e.g. "7abibi", "eshi" — Arabic sounds mapped to Latin letters and numerals) which remains common in mobile search in markets like Jordan, especially for younger and B2B audiences. We cross-reference Search Console's query report filtered to the Arabic locale against the English equivalent to see which intents are being served by the wrong-language URL, and we audit the page itself — is the Arabic content natively written, or does it read like a translation of the English draft.

The fix

Keyword targets are built around the root-and-pattern family, not a single literal phrase, so on-page content naturally covers the inflected forms real searchers use. Content is written independently in Arabic by someone thinking in Arabic — not drafted in English and translated sentence-by-sentence — because localization (adapting the message, examples, and phrasing for the reader) produces different, better-ranking copy than literal translation does. Metadata (title tags, meta descriptions, OG tags) is written natively in Arabic rather than transliterated from the English metadata, since search snippets are one of the first native-language signals a searcher evaluates. Reciprocal hreflang tags then make sure the Arabic page is actually served to Arabic-language searchers instead of the English URL outranking it by default.

Implementation details

On this site, RTL isn't a global stylesheet flip — every locale-aware component sets `dir={isRTL ? "rtl" : "ltr"}` explicitly (see components like the footer, table of contents, blog and case-study cards, and every Services page including this one), and layout is built with Tailwind's `rtl:`/`ltr:` logical variants (`rtl:mr-4 ltr:ml-4`, `rtl:right-4 ltr:left-4`) rather than fixed left/right values, so spacing and icon direction mirror correctly per-component instead of relying on a single blanket CSS override. Arabic URL slugs use literal Arabic directory segments — e.g. `/ar/خدماتنا/استشارات-تحسين-محركات-البحث` for this site's own SEO Consulting page — a pattern proven live across every Arabic service page on this domain (Schema Markup, Hreflang, Canonical Tags, JavaScript SEO, Core Web Vitals, and this page itself all resolve on literal Arabic slugs). Every static page's Arabic and English URLs are declared as reciprocal hreflang alternates from a single shared mapping (`page-mappings.ts`), so the two language versions never fall out of sync with each other.

Common mistakes

• Running an English keyword list through machine translation and treating the output as Arabic keyword research, instead of researching how Arabic speakers actually search for the concept.

• Machine-translating page content wholesale and publishing it as the Arabic page — it reads as translated, and "we write Arabic content independently rather than machine-translating" is a real technical decision with a real quality difference, not a marketing line.

• Applying a single dialect or register (usually formal MSA) to every page regardless of whether the query intent is informational (MSA reads naturally) or commercial/conversational (dialect often converts better).

• Building RTL with a single global `direction: rtl` override instead of per-component logical properties, which breaks icon mirroring and spacing in mixed-content or embedded components.

• Treating Jordan, Saudi Arabia, and the wider Gulf as one undifferentiated "Arabic market" with identical keyword and dialect assumptions, instead of researching each market's own search behavior.

Validation

We validate against Search Console filtered to the Arabic locale — confirming Arabic-intent queries are landing on the Arabic URL (not the English one), that impressions grow across the root-and-pattern keyword family rather than a single literal phrase, and that hreflang errors (checked via the Hreflang Implementation page's own validation process) stay at zero for the page pair. We also check click-through rate on the Arabic snippet specifically, since natively-written metadata is meant to read as native, not translated, in the search results.

Related service & solutions

Arabic SEO depends on the hreflang and technical foundations covered on these dedicated pages, and on the market-specific work covered on our Jordan and Saudi Arabia SEO pages.