What a crawlable blog post contains

A post that can rank is a document. It has its own URL, one title, a date in the page, and a canonical link. JavaScript can filter an index and switch a theme. The article text should not depend on it.

Write the page you would send to a teammate

Pick a problem you can explain with a concrete example: when to split a log, which token checks to test, why a list query should name its columns. A page written because a keyword tool reported volume, about a tool you have not run, does not help the person who arrived.

The tags that do the work

• a title and a description that state the point
• one h1, then sections
• a canonical URL
• a date in the HTML, not only inside a script
• links to the next post a reader would actually want
• a sitemap and an RSS feed of the same documents

If the script fails to load, the post is still the post. That is the same cut as keeping a serverless bundle small: the document is the product, and the tooling stays off the response.

What does not improve ranking

Publishing someone else's article under a new name does not. A thin page every morning, rewritten from whatever is trending, does not either. Search systems already see that pattern, and readers bounce. Frequency helps when each URL is one you would send to a teammate. A streak of pages that repeat a headline trains people to ignore the site.

Add a post when you can teach the rule without borrowing the page you learned it from.