SEO is a structural problem before it is a content problem
Most sites that cannot rank have a crawling problem disguised as a writing problem.
In short
- If a site does not rank, check crawlability before commissioning more writing.
- One canonical URL per page; parameter duplicates competing with each other is a self-inflicted wound.
- Speed is a ranking input measured on a phone, not on office fibre.
- Structure gets a page indexed. Only the writing makes anyone read it.
1
canonical URL per page, no duplicates
site rule
0
redirect chains tolerated
site rule
3
levels in the structure: home → section → article
site rule
When a site does not rank, the instinct is to write more. Usually the more useful first question is whether a search engine can reach what already exists, and whether the page it lands on is worth the crawl.
The four checks I run first
- Can it be crawled? A robots rule, a redirect chain or a JavaScript-only render can hide an entire catalogue.
- Is the structure unambiguous? One canonical URL per page. No parameter duplicates competing with each other.
- Is the page fast enough to be worth ranking? On mobile, on a mediocre connection, not on the office fibre.
- Does the markup describe the content? A title that says what the page is, a description that is not a keyword list, headings that nest.
What good structure looks like
A well-structured site tells a story without the words: home → section → article, each level linked by real URLs, each article pointing at related ones. Add structured data and a sitemap that reflects what actually exists, and the crawler does not have to guess.
Ranking is what happens when a crawler finds a page it understands, quickly, on a site that has earned the habit of being useful.
Where content still matters
Structure gets a page crawled and indexed. It does not make anyone read it. The writing still has to answer the question the visitor typed — and answer it before the fold, not after three paragraphs of preamble.
Fix the plumbing first. It is cheaper than a content calendar and it compounds.
Sources
Frequently asked questions
Is SEO a content problem or a technical problem?
It is structural first. A crawler that cannot reach or understand a page will not rank it however good the writing is; once crawling, canonicals and speed are sound, content decides position.
What should I check first when a page will not rank?
Four things, in order: can it be crawled, is there exactly one canonical URL for it, does it load in a couple of seconds on a phone, and does the title describe what the page actually answers.
Do I need structured data to rank?
No, but it makes a page easier to interpret and to quote. Article, FAQPage and BreadcrumbList markup is what lets search engines and AI answer engines lift a direct answer with attribution.
How many internal links should a post have?
Enough that a reader can continue: usually three to six, pointing at the section index and the genuinely related articles. Links added to hit a number read like links added to hit a number.