Keywords Optimized

I do not pick a keyword first. I start with the search intent behind the query. For a page targeting "best running shoes for flat feet", I open the SERP and look at what Google already shows. If the top results are buyer's guides with affiliate links, the intent is commercial, not informational. That tells me my page needs product recommendations, pricing, and a comparison table, not a general article about flat feet.

Once I have the intent, I run a proper [keyword research](/keyword-research/) session. I pull a list of related phrases from a tool and group them by topic. This is where [keyword clustering](/cluster-keyword/) comes in. For the running shoes example, I might cluster: "supportive running shoes flat feet", "arch support running shoes", "best stability shoes for overpronation". The primary keyword is the one with highest relevance and reasonable volume — typically "best running shoes for flat feet". The secondary keywords support the primary without competing for the same rank.

Next I decide where to place those terms. My hard rule: the primary keyword goes in the title tag, the H1, and the opening paragraph. But I never force it. If the sentence reads better with a synonym, I use the synonym. Google understands related language. I then spread the secondary keywords naturally across H2s, body paragraphs, image alt text, and the URL slug. One term per heading, no stacking.

I also check for [featured snippets](/featured-snippets/). If the SERP shows a paragraph snippet for a related question, I write a concise 40–50 word answer in a paragraph right after an H2. That has driven real traffic to my pages without extra effort. The final step is to review the page as a whole. If any keyword appears more than three times in a 500-word section, I rewrite to use a synonym or drop it.

This workflow is not fast, but it reduces waste. I would rather target one phrase well than ten phrases poorly.

The biggest trap is over-optimisation. I see it often: a writer tries to hit a 2% keyword density and the text becomes unreadable. Google's spam algorithm catches that. If you read your paragraph aloud and it sounds like a sales pitch, you have gone too far. A safer metric is to check that your target keyword appears in the title and once or twice in the body — nothing more.

Another watch-out is ignoring [search intent](/intent-seo/). I once spent weeks optimising a page for "SEO pricing" without noticing the SERP was full of software landing pages, not informational articles. My content flatlined. Now I check the SERP before writing and I ask: does the user want to buy, to learn, or to compare? If I cannot match that, I change the keyword or change the page format.

Do not forget about internal cannibalisation. If you have two pages both targeting "keyword optimisation", neither will rank well. Use a spreadsheet to map each URL to one primary keyword and a set of secondary terms. [Primary keywords](/primary-keywords/) should be unique across your site. If you find overlap, merge the pages or redirect one.

Also, watch out for keyword difficulty. A broad term like "SEO" has a difficulty score above 90 in most tools. You will not rank for it unless you are a major brand. Focus on [long-tail keywords](/long-tail-keywords/) with lower competition but clear intent. A phrase like "how to optimise keywords for ecommerce product pages" might have 50 searches a month, but it converts better because the user knows what they want.

Finally, never set and forget. Keywords lose relevance. Search behaviour shifts. I review my optimised pages every quarter and update the title, headings or examples if I see a drop in [keyword ranking](/keyword-ranking/).

I used to believe that the exact match keyword had to appear in the title tag or the page would never rank. I was wrong. Google has become much better at understanding context and related topics. I have seen pages rank in the top three for a phrase that never appeared in the title — the keyword was in the H2 and the paragraph instead. Now I still put the keyword in the title, but I do not panic if the phrasing needs to be adjusted for a clickable headline.

Another mistake: I underestimated the value of secondary keywords. Early in my career, I would target one phrase per page and ignore everything else. That left a lot of potential traffic on the table. When I started clustering terms and writing naturally, I saw a 30% increase in organic impressions for a set of product pages within two months. The extra coverage also helped the page rank for the primary keyword faster.

I also fell for the keyword density myth. A client once asked for exactly 2.5% density. I used a tool to count instances and it produced robotic copy. The page tanked after a Google update. Now I never measure density. I read the text and ask myself: would a human find this useful? If the answer is yes, the optimisation is likely fine.

Finally, I used to optimise the page first and check the SERP later. That is backwards. I now spend the first 30 minutes on keyword difficulty and SERP analysis before I write a single word. Saves hours of rewriting. If I had admitted that pattern earlier, I would have wasted less time on pages that never had a chance.