SEO: how search visibility actually works

What SEO actually is

Search engine optimisation is the work of making a site reachable, understandable and worth returning as an answer. Everything else is detail beneath those three.

Reachable means a crawler can get to the page and render its substance. Understandable means the page resolves to a clear topic and a clear entity. Worth returning means it answers the question better than the alternatives already ranking.

How search engines actually work

Three stages, and confusing them is the most common source of wasted effort. Crawling is discovery — can a bot reach the URL. Indexing is storage — is the page kept and understood. Ranking is selection — is it chosen for a given query.

A page can be crawled and not indexed. A page can be indexed and never rank. Diagnosing which stage is failing determines what work is worth doing, and it is the first thing any competent audit establishes.

What you can influence and what you cannot

You control your technical foundations, your content depth, your internal architecture and — indirectly — how third parties describe you. You do not control the algorithm, your competitors, or how quickly any of it takes effect.

This is why guaranteed rankings are a warning sign rather than a strong offer. A firm that promises an outcome outside its control is either misunderstanding the mechanism or hoping you will not notice.

Why it now has two surfaces

For twenty years being found meant earning a ranked link someone clicked. Generative engines changed the shape: the answer forms inside the result, and a growing share of searches end without a visit.

The foundations largely overlap — a site that cannot be crawled or whose entity does not resolve underperforms on both. What differs is measurement and the weight placed on independent corroboration.

On-page: what actually moves and what does not

On-page work is the most over-taught and least differentiating part of SEO. It matters, and it is table stakes rather than an advantage — every competent competitor has already done it.

The elements that still carry weight are the ones that clarify what a page is about and make it extractable. The ones that stopped mattering are the ones that were easiest to automate.

• Title element — still the strongest single on-page signal of topic. One page, one primary intent.
• H1 and heading hierarchy — structural, not decorative. A generative engine uses it to locate the passage that answers a question.
• Meta description — no ranking effect, real click-through effect. Write it as ad copy, not a summary.
• Internal links — the most underused lever on most sites. They pass authority toward the pages that need to convert.
• Image alt text — accessibility first, retrieval second. Both are reasons to write it properly.
• Keyword density — not a thing. The founding GEO study measured a negative effect for keyword stuffing.

Tactics we will not sell you

Some of the highest-volume searches in this category are for tactics that carry documented penalty risk. We publish this rather than compete for those terms, because ranking for them would attract exactly the wrong engagement.

• Bought links and private blog networks — a documented penalty risk, and the cheapest way to make a domain harder to recover than it was before.
• Publishing generated content at volume without review — two sites publishing roughly a thousand generated posts each were deindexed algorithmically, with no manual action.
• Doorway and template pages at scale — near-duplicate pages built for query permutations rather than for readers.
• Guaranteed rankings, guaranteed AI citations, guaranteed dates — nobody controls these, so nobody can guarantee them.