# AL MURIS INTERNATIONAL — 90-day launch and international SEO plan

Prepared 29 October 2026. Day 1 starts when the company approves its verified public information and authorises public launch on the intended domain. A private review site is not a Google-indexed launch.

## Operating goal

Generate qualified employer and agency enquiries, then measure which countries, sectors and pages produce useful commercial opportunities. Do not use raw traffic, candidate applications or unqualified WhatsApp clicks as evidence of recruitment contracts.

## Days 1–30: launch and establish reliable measurement

| Period | Work | Completion evidence |
|---|---|---|
| Days 1–7 | Verify legal identity, current licence, detailed address, actual services, destination authority, image permissions and any partners/deployments. Assign an enquiry owner and approve privacy/retention and responsible-recruitment policies. | Evidence register; empty unverified partner/deployment sections; signed content approval |
| Days 8–14 | Review English, Bengali and Arabic; test Arabic RTL, mobile forms, country prefill, source attribution, consent, upload protection, CRM permissions and confirmation state. Configure authorised transactional email and domain authentication. | Successful end-to-end test request; saved reference; received mail; protected CRM/file access; documented failures |
| Days 15–21 | Configure the production domain and HTTPS; verify Google Search Console and Bing Webmaster Tools; connect GA4 only with appropriate privacy choices. Submit the XML sitemap. Inspect representative pages in all languages. | Verified property ownership, sitemap acceptance and URL Inspection results; no accidental private/staging URLs in sitemap |
| Days 22–30 | Check canonicals, reciprocal hreflang, robots rules, crawlable navigation, broken links and page speed. Establish the first 28-day baseline. | Crawl report; mobile checks; baseline dashboard; list of indexing issues and owners |

Use language paths `/en/`, `/bn/` and `/ar/`. Every translated page should have a self-referencing canonical and reciprocal alternate-language links, including itself. Keep English as the `x-default` counterpart when it is the fallback. Country slugs identify subject matter; they do not mean the page is physically served from that country. Do not canonicalise Bengali and Arabic pages to English. Do not add unsupported languages to hreflang before their full pages exist.

Sources: [Google localized versions](https://developers.google.com/search/docs/specialty/international/localized-versions), [canonical guidance](https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls).

## Days 31–60: improve the pages with the best commercial fit

- Prioritise three to five markets only after route eligibility, company capacity and first enquiry evidence are known. South Korea and Hungary require the restricted-route treatment described in MARKET_RESEARCH.md; do not select them simply because a keyword looks attractive.
- For each priority market, create a useful employer briefing: relevant roles, exact skill questions, employer documents to review, accommodation/transport/shift considerations, actual process responsibilities and official route references.
- Add focused trade pages with evidence-based requirements. A welder brief should ask process/material/certification needs; a hotel brief should ask departments, language and shifts; an agricultural brief should ask crop/livestock tasks and season. Do not claim available candidates without confirmation.
- Obtain a native editor’s review of priority language pages. Validate local seed phrases using Search Console queries, Google Ads Keyword Planner if access is provided, and country-specific search-result review. Record the tool, country, language, date and match definition for every volume figure.
- Add genuine deployment stories only with employer/worker permission and verifiable facts. Improve links between market, sector, occupation and request pages.
- Review a weekly sample of rejected enquiries and form abandonment. Reduce unnecessary initial fields only if the CRM still receives sufficient information to qualify the order.

Completion target: a reviewed content brief and unique substantive page for each priority market; a source/date for every regulatory statement; one named owner for every open lead; no unsubstantiated public claims. These are operational targets, not promised rankings or lead volumes.

## Days 61–90: optimise from evidence

- Compare countries using qualified enquiries and progressed opportunities, not visitor counts alone. Compare both destination country selected in the form and source market page; they answer different questions.
- Examine country-page landing sessions, submitted forms, WhatsApp/email clicks and lead quality. Investigate gaps before spending on campaigns.
- Test one change at a time: shorter first-step enquiry, clearer trade description, relevant checklist or CTA placement. At low traffic, use observed customer feedback rather than declaring statistical winners.
- Improve pages with impressions but weak click-through when titles/descriptions do not match employer intent. Improve pages with traffic but weak enquiries after checking audience, eligibility and form errors.
- Consider a new language only when lawful route, employer demand, query evidence and an editor are available. Malay, Serbian, Italian, Japanese, Korean and Russian are candidates for evaluation, not automatically justified launches.
- Produce a day-90 review: published pages, index coverage, actual queries, qualified leads, pipeline by market, lead handling times, verified content gaps and next-quarter priorities.

## Event and reporting specification

| Metric | Definition and denominator | Source |
|---|---|---|
| Saved manpower enquiries | Unique server-saved manpower request IDs; deduplicate retries | CRM database |
| Saved partnership enquiries | Unique server-saved partner request IDs | CRM database |
| Qualified leads | Reviewed requests meeting defined employer identity, role/headcount, lawful route and commercial requirements | CRM qualification fields |
| Lead progression | Counts and rates moving New → Reviewed → Negotiating → Contract → Completed, with cancelled/rejected separate | CRM status history |
| Country performance | Requests by **destination country**, separately requests by **origin page market** | CRM fields |
| Occupation demand | Requests per role and requested worker counts, reported separately | CRM form fields |
| Form conversion | Saved requests / measured relevant landing sessions for the same period and consent scope | CRM plus analytics; label coverage limits |
| WhatsApp conversion signal | Clicks on official WhatsApp link; this does not prove a message was sent | Consented analytics event |
| Email conversion signal | Clicks on `mailto`; this does not prove an email was sent | Consented analytics event |
| Search visibility | Impressions, clicks, CTR and average position by query/page/country | Search Console aggregate reports |
| Response time | First substantive staff response timestamp minus request creation | CRM activity log |
| Data quality | Missing country/source, duplicate records, invalid email, spam and submission failures | CRM/operational review |

Use `form_start`, `form_submit_success`, `form_submit_error`, `whatsapp_click`, `email_click` and `partner_submit_success` only where implemented. `form_submit_success` must occur after confirmed server persistence. Do not put names, emails, phone numbers, document URLs or free-text requirement contents in GA4 event parameters or page URLs. Use coarse category codes and page paths.

UTM/referrer data is incomplete. Preserve landing page and attribution fields when consent/implementation permits. Distinguish “not available” from direct traffic where the origin is unknown. Search Console query data is aggregated and cannot reliably identify the exact query of a named lead. Do not join aggregate queries to personal CRM records as if this were observed attribution.

## Weekly scorecard and realistic targets

The first four weeks establish the baseline; there is no defensible fixed organic-lead forecast yet. Operational targets: every genuine lead has an owner; every form failure is logged and investigated; every published regulatory claim has a dated source; every public logo/photo has permission; every reported conversion uses a defined denominator. Set a response-time SLA that staff can actually meet, then measure median and 90th percentile against it.

Weekly report fields: reporting dates; total/qualified enquiries; country/role split; status movements; unresolved requests; measured sessions and consent coverage; WhatsApp/email clicks; Search Console clicks/impressions; important page changes; next three actions. Display zero or “no data” honestly.

Search visibility is not guaranteed. Google warns against largely duplicated regional doorway pages and keyword stuffing, and ignores the meta-keywords tag. Publish useful content and fix technical problems; do not buy ranking promises or manufacture backlinks.

Sources: [Google spam policies](https://developers.google.com/search/docs/essentials/spam-policies), [supported meta tags](https://developers.google.com/search/docs/crawling-indexing/special-tags), [how Google Search works](https://developers.google.com/search/docs/fundamentals/how-search-works).

## Access and evidence required for public launch

Domain/DNS or hosting access; verified legal entity/licence and actual service scope; authorised sender/email delivery provider and SPF/DKIM/DMARC configuration; Search Console/Bing verification; GA4 measurement ID if adopted; approved privacy/retention practices; administrator identities; real approved photography; partner permissions; deployment evidence; maintenance, backups and data recovery ownership. None of these should be represented as completed merely because integration code exists.