Server and Deployment SEO: Redirects, Status Codes, HTTPS and Canonicals
When a site goes live, the server becomes part of its SEO. The wrong redirect, a soft 404, a mixed http/https address or a crawler mistaken for a flood of traffic can quietly keep pages out of Google. Here is what to get right at deployment.

Most SEO advice stops at the page. But the moment a site is deployed on a server, the server itself starts answering Google — with status codes, redirects and headers — and those answers decide whether pages are kept, merged, or dropped from the index. A beautiful site with a careless server configuration will lose pages for reasons no content fix can touch.
This guide covers the server-and-deployment side of SEO: the settings that must be right when a site goes live, written for business owners and teams in Egypt, Saudi Arabia and Kuwait. It is the deployment half of technical SEO from development to launch, and it is where many "why won't this page index?" problems are actually solved.
Key takeaways
- The server must serve one canonical address: HTTPS, and one of www or non-www — never both.
- Status codes are messages to Google: 200 keep, 301 moved permanently, 404 gone, 5xx error.
- Use 301 (permanent) for moved pages; 302 (temporary) tells Google to keep the old URL.
- Avoid redirect chains and update internal links to point straight at the final URL.
- A soft 404 — an empty page that returns 200 — confuses Google; return a real 404 instead.
- If the server treats Googlebot as one heavy visitor and returns 503, pages drop out with "Couldn't fetch."
Why does the site need one address?
The same page can be reachable at several addresses: with and without www, over http and https, with and without a trailing slash. To a human they look identical; to Google they are different URLs competing with each other and splitting your signals. The fix is a server rule that forces one canonical address — on this site, https://nilexdigitalsystems.com — and permanently redirects every other variant to it. HTTPS is not optional: it is a ranking signal and a trust signal, and browsers now warn on sites without it.
What do status codes tell Google?
Every time the server answers a request it returns a status code. These are not technical trivia — they are instructions Google follows:
| Code | Meaning | SEO effect |
|---|---|---|
| 200 OK | Here is the page | Normal — keep and index |
| 301 Moved Permanently | This moved, forever, to here | Passes the old page's value to the new URL |
| 302 Found (temporary) | Temporarily elsewhere | Google keeps the old URL indexed — wrong for a real move |
| 404 Not Found | This page does not exist | Correct for removed pages; Google drops it |
| 410 Gone | Deliberately removed | A stronger, faster 404 |
| 5xx Server Error | The server failed | Repeated 5xx makes Google crawl less and drop pages |
How should redirects be done?
Use a 301 whenever a page permanently moves — a new URL, a changed structure, http to https, www to non-www. A 302 is for genuinely temporary cases; using it for a permanent move tells Google to keep the old address, which wastes the move. Two things ruin otherwise-correct redirects: chains (A redirects to B redirects to C — collapse them to A→C), and leaving internal links pointing at old URLs so every click takes a redirect hop. On a redesign, map every old URL to its new home before launch; we cover this in redesign without losing SEO.
What is a soft 404 and why does it hurt?
A soft 404 is a page that looks empty or like an error but still returns a 200 OK — an out-of-stock product with no content, a "no results" page, a half-built page. Google sees the 200 and tries to index nothing useful, then flags it. The fix is to make the page genuinely useful, or, if it is truly gone, return a real 404 or 410 so Google removes it cleanly instead of guessing.
Headers, private areas and the crawler-overload trap
Some pages should never be indexed — admin, accounts, checkout steps, staging. Keep them out with an X-Robots-Tag: noindex header (which works even for files like PDFs) rather than hoping robots.txt is enough. And watch one subtle deployment bug: if the server treats all crawler requests as coming from a single visitor and rate-limits them, it can answer Googlebot with 503 errors. Search Console then reports "Couldn't fetch" for the sitemap and pages, and indexing stalls — a real failure we have seen and design around by exempting legitimate internal and crawler traffic from visitor rate limits.
How Nilex helps
Nilex handles the server side of SEO at deployment: one forced HTTPS canonical address, permanent 301s for every alternate URL, correct status codes, real 404s instead of soft ones, X-Robots-Tag noindex on private areas, and crawl-friendly rate limits so Google is never blocked. We deploy the site, confirm Search Console can fetch the sitemap and pages, and watch the Page indexing report settle. It is part of our SEO-ready website service; the page-level groundwork is in technical SEO from development to launch and the result is every page indexed.
Frequently asked questions
Is HTTPS really necessary for SEO?
Yes. It is a ranking signal, browsers warn visitors without it, and some features require it. Every page should be served over HTTPS, with http redirected to it.
301 or 302 for a moved page?
301 for a permanent move — it passes the page's value to the new URL. Use 302 only when the change is genuinely temporary; otherwise Google keeps indexing the old address.
Why does Search Console say "Couldn't fetch"?
Usually the server returned an error (often 5xx or a rate-limit 503) when Google tried to read the page or sitemap. Fix the server response and the error clears on the next crawl.
Do I need to redirect old URLs after a redesign?
Yes — map every old URL to its new page with a 301 before launch, or you lose the ranking those pages had and send visitors to 404s.
Technical SEO From Development to Launch: Building a Site Google Can Actually Find11 November 2026 · 6 min
Get Every Page Indexed: A Clean Google Search Console With No Coverage Errors11 November 2026 · 6 min
Sitemaps, robots.txt and Structured Data: The Three Things That Get You Crawled, Indexed and Shown11 November 2026 · 5 min
