Technical SEO Consultant, Lebanon

What a crawler sees when the origin server is unreachable

When a Lebanese site goes offline because of a power cut, a connectivity problem or a struggling host, the damage to search depends on what the crawler receives in that moment. The worst outcomes are an error page served with a normal success status, which can be indexed as if it were the page's content, or a not found response, which tells search engines the page has gone. A short burst of server errors is far less harmful, and a temporary unavailable status with a retry hint tells crawlers to come back later.

Better still is for crawlers and visitors to receive a cached copy. A content delivery network can serve stale pages when the origin cannot answer, as long as caching is configured for HTML and not just images, and cache rules exclude carts, accounts and forms. For brochure sites, static generation of the main pages removes most of the dependency on a live server. Forms need a fallback too, such as a queue or a third-party form service, so inquiries are not lost while the origin is down.

I review Search Console's host status and crawl stats for spikes in errors, test how the site behaves with the origin blocked, and check DNS and certificate renewal, which can fail quietly during long outages. External uptime monitoring with alerts closes the loop. These recommendations go to your host or developer as specific configuration changes. The broader method sits on my technical SEO service page.

Declaring language and direction across three templates

A trilingual Lebanese site depends on every page stating its language and direction correctly in the HTML the server sends, not only after scripts run. The html element should carry the right language value, such as ar, fr or en, and the right direction. When that is set by JavaScript after load, crawlers that read the raw response may see an Arabic page labeled as English, and screen readers and browsers handle text badly.

Inside the page, mixed content needs care. A French product name inside an Arabic sentence, a phone number or a price in dollars can reorder itself unless the template isolates it. Titles and meta descriptions are especially exposed because search results display them without your styling. Components built for left-to-right layouts, such as carousels, breadcrumbs, arrows and progress steps, need mirrored versions for Arabic. Structured data should state the language of the page it describes, and names in markup should match the visible language.

Fonts are the performance side of the same issue. A Lebanese site may load an Arabic family and a Latin family with full French accents, in several weights, on every page. Subsetting each font to the characters it actually needs, loading only the weights in use and preloading the font behind the main heading can noticeably reduce render delay on weak connections. I check all of this template by template, in each language, and log the issues against the template rather than against thousands of URLs.

Online shops with shifting stock and buyers abroad

Lebanese retailers and producers selling online, including those shipping food, cosmetics, fashion and design pieces to customers abroad, often deal with stock that changes quickly. Imported items run out and return, prices are revised, and some products are available only for local delivery. If the site handles this inconsistently, search engines see products appearing and disappearing, prices in markup that no longer match the page and category pages full of items nobody can buy.

The fix is a written availability policy that the platform follows automatically. Temporarily unavailable products keep their page, stay linked from their category and show an accurate unavailable status in markup, ideally with a restock option. Discontinued products redirect to a close alternative or the category, or return a gone response if nothing relevant remains. Products sold only locally state that clearly, so visitors abroad are not misled. Availability and price in markup must come from the same source as the visible page, whether that is the store's database or an ERP or inventory feed.

Faceted navigation and variants complete the picture. Color, size and price filters, plus sort orders, can generate large numbers of near-duplicate URLs; a small set of filtered pages that match real searches can be indexable while the rest are kept out. Hosted shop platforms sometimes create several paths to the same product, so canonicals need checking. My ecommerce SEO page covers the wider approach.

Structured data for restaurant groups, hotels and clinics

Hospitality and health are significant parts of Lebanon's online search, and their sites benefit from precise structured data. A restaurant group can describe each outlet as its own restaurant entity with its address, cuisine, price range wording, accepted reservations and a link to a menu published as a real web page. Hotels and guesthouses can describe the property, its amenities and contact points. Clinics and medical centers can describe the organization, its departments and its practitioners, as long as every claim matches what the page shows and what your medical and legal advisors approve.

The organization itself should be defined once, with a stable identifier, its legal and trading names in Arabic, French and English, logo, contact points and links to verified profiles. Each outlet or branch then refers to it. This consistency matters when the same group operates venues in Lebanon and abroad, because search engines and AI assistants need to tell the Beirut location from a sister venue in Dubai or Paris.

Opening hours are the weak point. Seasonal schedules, holidays and unplanned closures change often, and markup, visible hours and business profiles drift apart. I recommend a single source for hours that feeds the page, the markup and, where possible, the business profiles. The modeling method is explained on my schema markup consulting page, and multi-outlet listings tie into my multi-location local SEO work.

Agency handovers, logs on shared hosting and consent tools

Many Lebanese sites are built by local agencies and freelance developers, some of whom also build for clients in the Gulf and Europe. The quality can be high, but handovers are often thin: nobody wrote down which plugin generates the sitemap, why a redirect exists or where the staging copy lives. When the developer moves on, the next change breaks something. I document the technical setup as part of the audit, so the next team inherits a page-type inventory, redirect rules and a release checklist.

Logs are another gap. Shared hosting plans frequently do not expose raw server logs, which leaves no record of how crawlers behave. If the site sits behind a CDN, its logs usually fill that gap, and even a short sample shows which URLs crawlers request and what they receive during outages. Redesigns and platform moves follow the redirect-map method on my SEO migration consulting page.

Consent tools and tags are worth reviewing too. Many Lebanese sites receive visitors from the European Union, whose rules can require consent before tracking, and Lebanese law may also apply; your legal advisor decides the model. I check that the banner does not shift the layout or block content and that tags load in a sensible order. Results are reported as indexing, crawl health and organic inquiries, never as promised rankings. The Lebanon international SEO page covers markets abroad, and my Lebanon SEO consulting page covers content and local search.