SEO

If I had to explain SEO in one sentence, it's this: help Google understand what your page is about, and help a person feel satisfied after reading it. Google doesn't care about your business — it cares about finding the best answer for a search. If your page is that answer, you'll rank. If it isn't, no amount of title-tag tinkering will get you to the top.

That sounds obvious, but I've seen people spend months adding keywords and building links to a page that nobody actually needs. A better starting point is a real question. Say someone searches "how to clean a cast iron pan without soap". If you write a page that gives a clear, step-by-step answer, with photos and a warning about what not to do, Google sees user engagement — people stay on your page, they don't bounce back to the SERP — and that signal is worth more than any keyword trick.

The fundamentals still hold: descriptive titles, clean URLs, fast load time, and links from other sites that trust you. But the engine behind it all is usefulness. Get that right and SEO becomes less of a guessing game.

SEO is not the right channel for every business. If you run a local plumbing company and a customer only ever finds you through a friend's recommendation or a Google Maps search, then a national organic ranking campaign is a waste of money. What matters there is a properly claimed Google Business Profile and proximity. The [Google Search](/google-search/) algorithm respects local intent differently.

For an e-commerce store selling niche products, SEO can be your primary traffic driver. I worked on one where 80% of new customers came from organic search. A single product guide — "best running shoes for overpronation" — brought in 2,000 visitors a month. That page needed a good title, solid internal links to product pages, and a few relevant backlinks. The tradeoff? It took six months to see meaningful traffic. Paid ads would have delivered clicks in hours, but at a cost per click that ate into margins.

A decision rule I use: if your customers search for a problem before they search for your brand, then SEO matters. If they only search for your brand name, you probably need a different strategy. The edge case is informational queries — people looking for answers, not products. Those are the easiest pages to rank because you can satisfy the intent fully without any sales pitch.

You see SEO every time you see a search result that isn't an ad. The blue link with the green URL and the snippet of text? That's the result of a title tag, a meta description, and a URL that tells both Google and the user what the page is about. The recipe with a photo and star rating? That's structured data — a vocabulary called Schema.org — that Google turned into a rich result.

I learned this the hard way when I added structured data to a simple guide on espresso ratios. The page started appearing with a calorie count and cooking time directly in the search result. Click-through rate went from 3% to 11%. The change took one afternoon of markup and a week for Google to recrawl.

Other visible bits: featured snippets (the boxed answer at the top of some results), "People also ask" dropdowns, and video thumbnails. All of them come from making your content structured and scannable. If you write a clear definition under an H2, you have a shot at being the snippet. If your page has a video, use a proper video sitemap. The technical work is invisible to the user, but the way the result looks is what decides whether they click your link or the one below it.

The biggest tradeoff is time versus reliability. SEO takes months to show compound results; paid ads show immediate traffic. If you have a short launch window, SEO is not your primary play. But if you can afford to wait, the cost per visitor drops to near zero after you rank.

Another tradeoff I still wrestle with: optimising for search vs. optimising for humans. Sometimes the phrase people use is awkward. "Ski boot size chart" is a fine title — it matches what people type. But a title like "Can you sharpen a serrated knife?" reads naturally and answers a common question. The decision rule: write the title for a human first, then check if it contains a reasonable match for the search term. If it doesn't, adjust the heading or the content, not the other way around.

There's also a risk of over-optimising. I once worked with a site that had a 400-word page stuffed with every synonym for "credit card rewards". It ranked well for two months, then a core update dropped it to page three. The content was thin and the links were bought. A better approach is to build a genuinely useful page that earns links naturally. That takes longer but the effect lasts longer too.

If you outsource, be careful. Agencies often pitch [SEO services](/seo-service/) that focus on quick wins like directory submissions and exact-match anchor text. Those tactics can backfire. I wrote more about what has and hasn't worked in my [SEO tips](/seo-tips/) page.

I spent my first year thinking SEO was a game of keywords. I'd cram "best coffee grinder for espresso" into the title, the H1, the body, and the alt text of every image. It worked — for a while. Then Google updated its handling of keyword stuffing (Panda, or maybe Penguin — I wasn't paying close attention to the name) and my traffic halved overnight. The pages were useless to anyone who actually knew about coffee.

I also ignored mobile. In 2020 I looked at analytics for a site where 68% of sessions came from phones. The text was 12 pixels, the buttons overlapped, and the menu required two taps. Google's mobile-first indexing meant that even if a page had great content, a poor mobile experience hurt its desktop ranking too. Fixing that gave me a 30% lift in organic traffic within a month.

The third mistake: I thought link building was a numbers game. I traded links with anyone who asked. Eventually I had links from spammy directories that hurt more than helped. Now I only pursue a link if the referring site is relevant and I'd be happy for my readers to see it there. That's a slower process, but the [SEO basics](/seo-basics/) guide I wrote later taught me that trust is earned, not bargained for.

I still get things wrong. I'll launch a page, optimise the schema, build a few good links, and get impatient when it doesn't rank in six weeks. Patience is the part of [learn SEO](/learn-seo/) that nobody mentions.