SEO Process

I start every SEO project by looking at the search engine results page for the target query — not a keyword tool. For example, on a recent project for "best running shoes", the SERP was dominated by round-up articles and YouTube videos. That told me the search intent is commercial investigation with a preference for visual comparison. I didn't need a long-form guide; I needed a well-structured product comparison with embedded video.

Then I audit the technical state: is the page even indexed? I check via site: search and Google Search Console. If not, I fix crawl issues. On a client site last quarter, three high-potential product pages were blocked by a misconfigured robots.txt. That took ten minutes to fix and brought a 40% increase in indexed pages within two weeks.

Next, on-page: I align the title tag and H1 with the query, but I avoid stuffing. I write a meta description that sells the click — not a summary. I also review internal links. In that running shoes project, the client had a dozen product pages with no links from the blog. I added contextual links from a popular "how to choose" article. That doubled indexed pages in two months.

Finally, I measure. I use Search Console to watch click-through rate and impression position. If rankings rise but clicks don't, the title or snippet needs work. That's the loop: SERP → technical → on-page → links → measure. I repeat it weekly for high-priority queries and monthly for the rest. For a full list of checks, see my [SEO checklist](/seo-checklist/).

Three traps I see often.

First, treating SEO as a one-time setup. I once optimised a page for "SEO services" and saw good rankings for six months. Then Google updated its review algorithm, and the page tanked because we hadn't refreshed content or added new signals. SEO needs monthly attention — at minimum, a content refresh and a link-health check.

Second, ignoring search intent for the sake of keywords. A client wanted to rank for "digital marketing agency" but every SERP result was a service page. They had a blog post. That's an intent mismatch. I redirected the effort to a new service page and aligned the [SEO strategy](/seo-strategy/) around the commercial SERP format.

Third, technical debt. I often find sites with dozens of 404s, slow-loading images, and broken internal links. Google de-prioritises crawl on these sites. My rule: fix all 4xx and 5xx errors before doing any new content. Also watch out for duplicate content from faceted navigation — use canonical tags or noindex. I once saw a site with 300 indexed URLs that were all filter variants of one page. That diluted the ranking signal for the canonical product page.

Another watch-out: measuring vanity metrics like total keywords ranking without looking at traffic or conversions. I had a site in the fitness niche ranking for 2000 keywords but getting only 50 visits because they were all long-tail informational queries. Focus on queries that match your business goals. [Organic SEO](/organic-seo/) is about quality traffic, not keyword count.

Finally, don't chase every ranking position equally. Positions 1-3 capture nearly 60% of clicks; the drop-off after position 5 is brutal. Prioritise pages that are already on page one — small tweaks there yield bigger gains than pushing page-two content.

I used to believe that the more keywords I stuffed into a page, the better. My first SEO attempt wrote a 3 000-word article repeating "SEO tips" five times per section. It ranked nowhere. Google's [semantic search](/semantic-search/) understanding means one well-placed term and its synonyms work better. Now I write for topic coverage, not keyword density.

I also ignored internal linking for years. I'd build links from external sites but leave pages orphaned. That's a mistake: internal links are the easiest authority you can pass. Now I ensure every new page has at least three links from existing content. On one site, that simple change lifted the average [SEO ranking](/seo-ranking/) of new pages by 15 positions within three months.

Another error: I thought content length was the ranking factor. I spent weeks producing 5 000-word guides for "SEO for beginners". A competitor's short listicle outranked me because it better matched the quick-answer intent. I now check the word count of the top three results and aim for the same range — not a single arbitrary number.

I also chased algorithm updates too early. After the Helpful Content Update, I deleted pages that were technically useful but not "helpful" enough. That was rash. I should have improved them instead. Today I judge content by whether it answers the user's next question better than the top result. That shift changed my process entirely.