Before You Write Another Blog Post, Fix These Technical Basics

Every few weeks someone asks me to look at a site that has stopped ranking, and the assumption is always the same: we need more content. Usually we do not. Usually there are eight or ten existing pages that a search engine cannot read properly, and adding a ninth does not fix that.

So I start at the bottom. Can Google fetch the page at all? I check robots.txt, then the rendered HTML rather than the source, because a lot of modern builds ship an empty shell and fill it in with JavaScript. If the content only exists after hydration, you are asking the crawler to do extra work and hoping it bothers.

Next, is each page about one thing? A page trying to rank for web design, SEO, ads and ecommerce at once ranks for none of them well. I map pages to the searches they should own, then look for the ones that overlap. Two pages chasing the same query compete with each other, and the fix is usually to merge them rather than write a third.

Then metadata, which is the cheapest thing on this list and the most commonly neglected. A title should say what the page is and where you are, not just repeat the brand. A title tag is often the highest-leverage line on a page and usually the last one anyone updates.

Structured data comes after that. It will not rank a page on its own, but it changes how the result looks, and better-looking results get clicked. Organization, WebSite and FAQPage cover most business sites, and a service or product page should describe the actual thing it sells.

Finally, speed. Not the number in a lab test, the number real visitors get. I look at field data for Core Web Vitals, find the slowest page that also gets traffic, and fix that one first instead of chasing a perfect score across the whole site.

The order matters. Content is the last lever, not the first, because content on a site that cannot be crawled or understood just adds to the pile.