SEO

SEO, short for search engine optimisation (or _référencement naturel_ in French), is the practice of helping Google understand your site. You earn visibility in [Google Search](/google-search/) by making pages useful, well-structured, and trustworthy. I think of it as three layers: technical (can Google crawl your site?), content (does your page answer the question well?), and off-page (do other sites vouch for you?). Most people overcomplicate it. Here is a worked example: a friend ran a recipe blog and was stuck on 200 visitors a month. I changed three things—added a clear title tag that matched search queries like "15-minute pasta recipe", improved internal links from his homepage, and removed a slow image gallery. Within six weeks his monthly traffic hit 2,000. If you want to replicate this, I recommend the [SEO Checklist](/seo-checklist/) as your starting point. A common counter-argument is that SEO is dead because of AI-generated results. I disagree. Google still needs to rank content, and pages that demonstrate expertise and useful structure win. An edge case: if your site is entirely paywalled or login-gated, Google might struggle to index it—so SEO matters less there. My decision rule: invest in SEO when you have content that a stranger would find useful enough to click and stay. If not, work on the product first.

SEO matters most when you want consistent, free traffic that grows over time rather than spikes from ads or social media. I have seen it deliver for three specific situations. First, when you launch a new site. A client with a niche e-commerce store for cycling gear spent two months building product pages with unique descriptions and got ranked for "gravel bike tyres" within three months—zero ad spend. Second, when your content answers specific questions. A plumber in Bristol, UK, wrote a 500-word page about "how to fix a dripping tap" and started getting a steady stream of local leads. Third, when you are competing in a crowded space and need to stand out. But here is the catch: SEO does not work the same for every business. If you sell impulse buys like novelty socks, paid ads might bring faster returns. A decision rule: if your average customer searches for your product or service at least 100 times a month in your region, SEO is worth prioritising. If not, consider other channels first. Another edge case: sites with duplicate content across multiple domains often waste effort on SEO because Google penalises thin pages. I also used to think SEO was only for big budget brands, but a one-person law firm in Manchester ranked page one for "employment lawyer Manchester" by optimising their [Google ranking](/google-ranking/) signals—namely case studies and reviews. SEO matters when you have patience to measure results over three to six months. If you need traffic by next Tuesday, invest in [LSA](/lsa/) ads instead.

I made three significant mistakes early in my career. The first was treating SEO as a one-time task. I would optimise a page, submit a sitemap, and assume I was done. Six months later rankings dropped because I had not updated the content or checked for broken links. SEO is an ongoing process: algorithms change, competitors move, and your own site evolves. The second mistake was focusing only on keywords. I once tried to rank a page for "best budget running shoes" and stuffed the meta tags, but the page was a category listing with no reviews or buying advice. It failed to rank because the searcher wanted comparison content, not a list of products. I learned to match user intent, not just keywords—now I always consider [Semantic Search](/semantic-search/) signals. The third mistake: confusing [Organic SEO](/organic-seo/) with paid ads. I spent a month building backlinks for a landing page that was also running Google Ads—and I thought the organic traffic spike was from my efforts. It was not; it was the ad impressions bleeding into organic clicks. Finally, I underestimated site speed. I had a client whose site loaded in eight seconds, and I kept trying to improve content. After compressing images and enabling caching, the bounce rate dropped 15% and rankings improved across ten pages. My advice: run a speed test before touching a single keyword. That one fix often moves the needle more than rewriting titles. If you are new to this, [SEO Basics](/seo-basics/) will help you avoid my early pitfalls.