SEO works in three stages: (1) crawling — Googlebot finds and reads your pages via internal and external links; (2) indexing — Google stores parsed content in its index alongside metadata about quality, topic, and authority; (3) ranking — for each query, Google's ranking algorithm scores indexed pages on relevance, intent match, authority, freshness, user-experience signals, and now AI-summary suitability. Pages ranked 1–3 capture 60%+ of organic clicks (less when AI Overview present).
Googlebot starts at known URLs (your homepage, sitemap, links from other sites that already link to you) and follows links to discover new pages. Pages without internal links don't get discovered. Pages blocked by robots.txt or noindex tags don't get indexed even if discovered.
You help crawling by: submitting a clear XML sitemap, using descriptive internal links, ensuring fast page speed (Googlebot has a crawl-budget per site), avoiding orphan pages.
Once crawled, Google parses the page (HTML, schema, content) and stores it in the index. The index includes: textual content, structured data (schema), image and video metadata, link relationships, calculated quality signals (E-E-A-T), and now AI-suitability scores.
You help indexing by: using semantic HTML (h1, h2, p, ul, ol, etc.), adding schema markup (Organization, Article, FAQPage, HowTo), keeping URLs human-readable, ensuring content is unique (not duplicated from other sites or thin AI-generated).
For each query, Google scores indexed pages on multiple factors. The publicly known factors include: relevance (does the content match the query intent?), authority (do quality sites link to this page?), freshness (is the content current for time-sensitive queries?), user-experience (Core Web Vitals, mobile-friendly, no intrusive interstitials), and increasingly AI-summary suitability (does this page provide a clear, structured answer that AI Overviews can quote?).
Search intent matters most. Pages ranking for 'how to fix a leaky tap' are educational; pages ranking for 'plumber london' are commercial. Mismatched intent ranks badly.
Google's AI Overview feature now appears on 25–35% of commercial queries. It synthesises an answer from multiple ranked pages, citing each source. CTR on rank-1 organic drops 30% when AI Overview is present.
Optimisation shift: get cited, not just ranked. Citation factors: lead-with-direct-answer in first 50–80 words, structured data (FAQPage, HowTo), primary-source citations within content, date-stamped content with regular updates, structured lists (numbered or bulleted) that AI parses preferentially.
Backlinks (links from other sites to yours) signal authority. Quality matters: one editorial link from a sector publication beats 50 directory links. Google's PageRank algorithm (modernised, but the core logic remains) treats links as votes weighted by source authority.
How to get links: digital PR (data studies, press releases, expert quotes), guest posts on sector sites, resource-page outreach, podcast appearances, broken-link building, brand-mention reclamation. Avoid: buying links, PBN networks, comment-spam, low-quality directory submissions.
"Modern SEO is a content + link + UX trifecta. Strong content with no links ranks slowly. Strong links with weak content rank but don't convert. Strong UX without content has nothing to rank."
Three reasons: (1) Google's crawl-and-index cycle is gradual — new pages can take 4–12 weeks to appear in results; (2) authority builds via links and topical depth, both of which compound over months; (3) algorithm trust signals (consistent publishing, low bounce, repeat visits) take time to accrue.
The compounding effect is the moat. After 12–18 months of consistent SEO work, organic traffic typically accelerates non-linearly — this is why patient SEO investment beats sporadic bursts of work.