Top 5 Technical SEO basics for Web Developers

A practical primer on technical SEO considerations developers should keep in mind while building websites.

Crawlability comes before everything else

Search visibility starts with a page a crawler can reach and read. Before anyone argues about keywords, it is worth confirming that the routes you have built are actually linked from somewhere, that robots.txt is not quietly excluding a whole directory, and that the sitemap you generate matches the pages that really exist.

The common failures are dull rather than exotic: a staging robots file shipped to production, an orphaned section reachable only from a menu that renders after a script runs, or a sitemap left over from a previous structure. All three are the developer's to fix, and all three are cheap to check.

Give every page a single, honest structure

One h1 per page, headings that descend in order, and a title element that describes the page rather than the site. That is most of on-page structure, and it is the part a developer controls directly.

The same applies to meta descriptions. They are not a ranking lever, but they are the sentence a person reads before deciding whether to click, so writing them per page rather than templating one across the site is worth the extra few minutes.

Redirects and error handling

When a URL changes, a permanent redirect keeps the link equity and, more importantly, keeps the people who bookmarked it. Chains of redirects that pass through three intermediate URLs are slower and lose more than a single hop, so it is worth flattening them whenever a structure is reorganised.

Equally important is what happens when a URL genuinely does not exist. A real 404 response with a useful page beats a soft 404 that returns the homepage with a 200 status, which quietly manufactures duplicate content out of every mistyped address.

Performance is a technical SEO problem

Page weight, render-blocking resources and unoptimised images are development concerns that surface as search concerns. Compressing images properly, deferring what does not need to run first, and keeping the critical path small usually produces a bigger measurable improvement than any amount of copy tuning.

Serving images at the size they are displayed, rather than shipping a 2000px file into a 600px slot, remains the single most common easy win on a content site.

Mark up what the page actually is

Structured data lets a page state plainly what kind of thing it describes. For an article-driven site, marking up the article, its publication date and its publisher costs very little and makes the page legible to machines in a way plain HTML does not.

None of these five points is glamorous. Together they are the difference between a site that search engines can work with and one that fights them by accident.