How to write a page that ranks

The advice here has been the same for a decade and is still mostly ignored, because it is less fun than the alternatives. A page ranks when it answers a specific question better than the pages currently answering it. Everything below is in service of that sentence.

One page, one job

Decide the single search this page is for and write it at the top of your draft. If you find yourself covering a second subject because it is related, that is a second page. Two half-covered subjects on one page lose to two pages that each cover one.

This also fixes a problem people create for themselves, where several of their own pages target the same phrase and none of them wins. Pick the page that should own it, and make the others link to it instead of competing with it.

When your own pages compete

This is worth doing properly because it is common and invisible until you look. Two or three pages all half-targeting the same phrase split the signal between them, and the engine picks whichever it likes least.

1. 01Search your own site for the phrase with the site: prefix and see how many of your pages come back for it.
2. 02In Search Console, check which page actually ranks for the query. It is often not the one you intended.
3. 03Choose the page that should own it, on the basis of which one best matches the intent, not which one you prefer.
4. 04Make the others point at it with descriptive link text, and cut the sections where they were half-answering the same question.
5. 05If two pages are genuinely the same page, merge them and redirect one to the other rather than leaving both thin.

Structure the page around the follow-up questions

Answer the main question near the top, in plain words, before any preamble. Then use your headings for the questions a reader will have immediately afterwards. You already know what those are: they are what people asked you when you explained this in person.

This is also how you earn the related searches around the main one, without writing a separate page for each. A section that genuinely answers a sub-question can rank for that sub-question on its own.

The shape that works, in order: the answer in two or three sentences, then why it matters or when it applies, then the sections answering the obvious follow-ups, then what to do next. Notice that the background and the history go late or not at all. Most pages invert this and lose the reader before the answer.

Depth, not length

Long pages often rank, which has been widely misread as long pages rank. What actually happens is that thorough pages tend to be longer. Padding a page to reach a word count adds the one thing you do not want: text a reader skims past looking for the answer.

A better test than word count: could somebody who read your page explain the subject to a friend without opening anything else? If yes, the page is long enough. If no, it is short, whatever its length.

A second test, which catches padding specifically: read each paragraph and ask what a reader would lose if it were deleted. If the answer is nothing, delete it. Pages get better at this stage by subtraction more often than by addition.

Titles and descriptions

The title is the one on-page element still doing real ranking work, and it is also an advert competing with nine other adverts on the results page. Make it say exactly what the page is, include the phrase naturally, and put the distinctive part first because the end can be cut off.

• Write the title for the person deciding which result to click, not for a keyword counter.
• Do not repeat your brand name at the front of every title. It pushes the useful words out of view.
• The meta description does not affect ranking, but it affects whether anyone clicks. Write it as the reason to choose this result.
• If you leave the description blank, a snippet from the page is used instead, which is sometimes better than a bad one you wrote.
• Check what is actually being shown. Search engines rewrite titles when they think yours fits the query badly, and seeing the rewrite tells you what they think the page is for.

Updating beats publishing

The highest return activity in small-site SEO is not writing the next page. It is fixing the pages that already rank somewhere on page two, because they have already proved the engine will consider them.

1. 01Find the pages ranking between roughly position eight and twenty for something you care about.
2. 02Search that query and read the three pages above you properly. Write down what each covers that you do not.
3. 03Add those sections if they belong, and cut whatever is now wrong or dated.
4. 04Improve the title so it matches the query more exactly than it did.
5. 05Add two internal links to the page from older pages that are relevant.

That is an afternoon per page, and it moves pages that are already close. Writing a new page from nothing is several days and starts at zero. Most sites have the better option sitting untouched.

The mistakes that cost the most

• Three hundred words of introduction before the answer. People leave, and that is measurable.
• Writing for a search you did not check the results for, so the page is the wrong format.
• Publishing twelve thin pages on near-identical subjects instead of one good one.
• Copying the structure of whoever ranks first, which produces a page with no reason to outrank them.
• Never updating. A page that was right two years ago and is wrong now will slide, and fixing it is cheaper than writing a new one.