There's a lot of bad advice about SEO content out there. Keyword stuffing. 2,000-word minimums. AI-generated pages by the hundreds. I've tried most of it — and the content that actually ranks and earns traffic follows a different, quieter set of rules. This is the playbook I use for my own sites and for clients. It's not glamorous. It works.

Rule 1: Start with intent, not keywords

Before writing anything, I ask: what is this person actually trying to do? Someone searching "how to start a blog" wants a guide. Someone searching "best blogging platform" wants a comparison. Someone searching "wordpress vs blogger" wants a decision. The same keyword can hide completely different intents.

Search intent decides your entire article — its structure, its length, even its tone. Write for the intent, and Google's algorithm (which is basically a giant intent-matching machine) will reward you.

Rule 2: Build topic clusters, not random posts

One-off articles are lottery tickets. Clusters are compounding investments. Pick a pillar topic — say, "content marketing" — and write one comprehensive pillar page, then surround it with supporting articles: "content calendar template", "how to write a blog post outline", "content repurposing guide". Every supporting article links to the pillar; the pillar links back out.

Why this works: search engines see a site that clearly owns a topic. Every new article makes the whole cluster stronger. That's how small sites outrank big ones — depth beats breadth.

My ratio: for every 1 pillar page, publish 5–8 supporting articles over time. Each one is short-term work that compounds long-term authority.

Rule 3: Structure for the reader and the snippet

Most people who land on your page won't read it — they'll scan it. Structure your article so scanning answers their question:

Rule 4: Write longer only when it's genuinely better

Ignore the "2,000+ words minimum" myth. The right length is the length that fully answers the intent — nothing more. A great 800-word answer beats a padded 2,500-word one, and readers (and Google) know the difference. When my articles do run long, it's because they cover real subtopics, not because I'm chasing a word count.

Rule 5: Titles and meta are part of the copy

Your title is a promise; your meta description is the ad. I spend real time on both:

Rule 6: Use AI for research and speed — never for autopilot

AI has made content production 10x faster, which is exactly why most AI content ranks nowhere: everyone published the same generic paragraphs. My workflow uses AI as a research and drafting engine, then adds what AI can't: real examples, honest opinions, original data and my own voice.

  1. Research: AI summarises top-ranking pages and surfaces gaps — what they cover, what they miss.
  2. Outline: structure based on intent, not on what AI would write.
  3. Draft: AI generates a first pass; I rewrite sections with my experience.
  4. Optimise: AI checks headings, keywords and readability; I keep the final call.
  5. Publish & update: content ages. I schedule regular refreshes — Google rewards fresh, accurate pages.
“Google doesn't reward content that exists. It rewards content that answers — better than everything else on the page one.”

The 90-day reality check

Set expectations early: new content rarely ranks overnight. Google needs time to crawl, index and test your pages against searchers. My typical timeline looks like this:

If a page isn't moving after 3–4 months, I diagnose: weak intent match, thin content, bad internal links, or a topic with no realistic search volume. Fix the diagnosis, not the symptoms.

The bottom line

SEO content is boring in the best way: intent research, honest writing, clean structure, consistent publishing. There's no hack that outranks that. The sites that win aren't the ones with the most content — they're the ones with the most answer-per-word. Write like that, publish consistently, and the rankings follow.

← Previous: AI in Research