Every week someone sends us a blog that “should be ranking.” It has the keyword in the title. It is 2,000 words long. It was published four months ago. And it sits on page four, getting eleven visits a month, most of them from the person who wrote it.
After seven years and a lot of these conversations, the reasons are almost always the same. Not one of them is word count, and not one of them is keyword density. Here is what actually goes wrong, and what the articles that do rank have in common.
1. Nobody checked what already ranks
This is the big one. Before writing a word, open an incognito window and search the exact phrase you want to rank for. Look at the top five results. What format are they? How long? What do they cover that you were not planning to?
If the top results are all step-by-step guides and you wrote an opinion piece, you will not rank, no matter how good the opinion is. Google has already decided what people searching that phrase want to see. Your job is to give them a better version of that, not a different thing.
2. The article answers a different question than the search
“Best CRM for small business” and “what is a CRM” look like the same topic. They are not. The first person is ready to compare and buy. The second has never used one. An article that explains what a CRM is will never rank for the first phrase, because it wastes the reader’s time, and Google can see people bouncing back to the results.
Every keyword has an intent behind it: learn something, compare options, find a specific page, or buy. Match the intent and half the battle is done. Miss it and nothing else matters.
3. It was padded to hit a number
Somewhere along the way “long-form content ranks better” turned into “write 2,000 words no matter what.” So articles open with three paragraphs of history, restate the question, and repeat the same point in different words until the counter says enough.
Readers scroll past that. Google measures the scrolling. The articles that rank are as long as the topic needs and not a sentence longer. Sometimes that is 3,000 words. Often it is 900.
Word count is a result of covering the topic properly. It is not a target.
4. The structure fights the reader
Look at any page-one article and you will see the same skeleton: a short answer near the top, clear subheadings that each answer one question, short paragraphs, and lists where a list makes sense. This is not a style preference. It is how people read on phones, and it is how Google pulls featured snippets and how AI search tools pull citations.
The articles stuck on page four usually have the answer buried in paragraph nine, under a heading like “Introduction.”
- Put the direct answer in the first 100 words.
- Make every H2 a question someone would actually type.
- Keep paragraphs to three or four lines on a phone.
- Use a list whenever you find yourself writing “first… second… third…” in prose.
5. There is nothing in it that only you could have written
This is the one that has changed most in the last two years. Google’s guidelines now talk about experience, expertise, authority, and trust. In practice it means: does this article contain anything the writer actually knows, or is it a rewrite of the other results?
A real example, a number from a client project, a mistake you made and fixed, a screenshot of your own data. These are the things a generic article cannot have. They are also the things AI search engines look for when deciding which page to cite.
6. It was published and abandoned
No internal links pointing to it. No update after six months. No links from the rest of the site’s content. An article with no connections is a page Google has no reason to trust.
The pieces that climb get linked from three or four related articles, get a small refresh when something changes, and sit inside a cluster of content on the same topic. Topical authority is built, not written.
What the articles that rank have in common
Put the six points together and the pattern is clear. Ranking articles were researched before they were written. They match the intent of the search. They are exactly as long as they need to be. They are structured for scanning. They contain something only the writer could know. And they are connected to the rest of the site.
None of that is a trick. It is just work that happens before the writing, which is why most blogs skip it. If you want to check your own content against this list, the SEO and blog writing page explains how we run each step.
