Content and SEO

Search engine optimisation has a reputation for being a technical discipline of keyword density, backlink schemes, and metadata manipulation. That reputation is partly historical: early SEO was genuinely about gaming algorithms rather than serving readers. Modern search engine algorithms are more sophisticated, and the gap between what helps readers and what helps rankings has narrowed significantly. Good content is good SEO, with a small number of structural considerations applied on top.

Search intent and content fit

The most important SEO concept is search intent: the underlying goal behind a search query. A reader who searches “how to centre a div” wants a code example, not a history of CSS. A reader who searches “best CSS frameworks 2026” wants a comparison, not a tutorial. A reader who searches “Tailwind CSS” might want documentation, a tutorial, or an overview — the query is ambiguous.

Content that matches search intent ranks better than content that stuffs in keywords but does not answer the underlying question. This means the first step in SEO-informed content planning is understanding what the reader is actually trying to accomplish when they use the search terms you want to rank for — and then writing content that accomplishes it for them.

The content hierarchy for SEO

Search engines parse content in the same order that readers do: title first, headings second, body text third. The structural hierarchy of SEO-optimised content is:

Page title (H1) — should contain the primary search term and match what the reader expects to find. “Typography for the Web: Line Height, Measure, and Scale” signals clearly to both readers and search engines what the page is about.

Headings (H2, H3) — should answer related questions that a reader who found the page might also have. H2s serve as the sub-questions that the article addresses. A page about line height might have H2s on “What is line height?”, “How to set line height in CSS”, and “Line height for different devices”.

Body text — should use the reader’s language naturally, not insert keywords artificially. Keyword stuffing — repeating a term many times, unnaturally — is penalised by search algorithms and is unpleasant to read.

Meta description — not a ranking factor directly, but affects click-through rate in search results. A well-written meta description is a pitch to the reader: it should tell them specifically what the page covers and why it is the best result for their query.

Long-tail content and specificity

Generic, broad topics are extremely competitive to rank for. “Design” is effectively impossible for a new site to rank for. “Grid systems in CSS” is less competitive. “12-column grid vs fluid grid: when to use each” is less competitive still, and more useful to a reader with that specific question.

This is the long-tail content strategy: write specifically about narrow topics, answer those questions completely, and accumulate a large collection of specific articles rather than competing for a small number of broad terms. Each specific article that ranks for a narrow query brings a highly motivated reader — someone looking for exactly the answer the article provides.

Internal linking as a content practice

Internal links — links from one page on a site to another — are both an SEO signal and a reader service. They distribute authority across the site and create navigational paths between related content. Writing internal links naturally, within the text, as part of the recommendation to the reader (“read next: how people read”) is both better SEO than a generic “related articles” block and more useful to the reader.

The key discipline: link to the most relevant page, with anchor text that describes the destination accurately. Generic anchor text (“click here,” “read more”) neither helps search engines understand what the destination page is about nor helps the reader decide whether to follow the link.

Measuring content performance

SEO without measurement is assumption. The metrics that matter for content performance differ by content type:

Search performance — Google Search Console reports impressions (how often a page appears in results), clicks, and average position per query. An article with high impressions but low click-through rate has a positioning or meta description problem. An article with high click-through rate but high bounce rate has a content-matching problem — it attracts the right search intent but does not deliver what the reader expected. Both are actionable diagnoses.

Engagement — time on page and scroll depth indicate whether readers are engaging with the content or leaving immediately. A high-traffic article with a 20% scroll depth is effectively unread. The cause is usually a content-expectation mismatch (the visitor’s intent is not served) or a structural problem (the content does not give the reader a reason to continue past the first screen).

Ranking velocity — new content typically takes 3–6 months to reach its stable ranking position. Making judgements about content performance within the first 60 days is usually premature. Track ranking positions at 90-day intervals for new content; at 30-day intervals for updated content.

Content decay — most articles lose ranking position over time as competitors update their content and algorithms change. Systematically reviewing content older than 12 months — updating examples, correcting outdated claims, expanding thin sections — extends the useful life of existing articles. Updating the dateModified schema property accurately signals freshness to search engines.

Structured data and schema markup

Structured data — most commonly implemented as JSON-LD embedded in a page’s <head> — provides machine-readable signals that search engines use to generate rich results: article author and publication date, breadcrumb trails, FAQ accordions, product prices, star ratings, event dates. These do not affect the page’s ranking directly, but they affect how the page appears in search results, which affects click-through rate.

The most useful schema types for editorial and documentation sites:

Article — marks up the article headline, description, author, datePublished, and dateModified. Google uses dateModified to assess freshness; keeping this accurate matters for content that is actively maintained.

BreadcrumbList — allows search results to show the page’s structural path (“Design Hubs > Typography > Type Scale”) rather than the raw URL. Implemented on every page in a multi-level hierarchy.

FAQPage — if an article contains a question-and-answer structure, the FAQ schema enables Google to surface individual Q&As in search results, expanding your visual footprint beyond a single link.

The specification for all schema types is at schema.org. Validate structured data with Google’s Rich Results Test before publishing. Common mistakes: forgetting to update dateModified on content revisions; omitting required properties (author for Article, name and item for BreadcrumbList); using incorrect property names (the spec uses camelCase strictly).

The final section of this hub covers voice and tone — starting with voice vs tone, the distinction that keeps brand language consistent without making it rigid.