SEO tips that survive contact with a real site

Why every tip here is written as a mistake

Generic advice is unfalsifiable. 'Create quality content' is true and useless, because nobody sets out to create bad content. It gives you nothing to check, nothing to fix, and no way of knowing whether you followed it.

So each rule below is written the other way round: as the specific mistake it stops you making, and what that mistake costs you in practice. If you can't name the failure a tip prevents, the tip is filler.

Fifteen rules, four groups. Then five popular tips that don't survive scrutiny, because unlearning is usually cheaper than learning.

Tips 1–4: get indexed before you get clever

Nothing else on this page matters if Google can't see the pages. This is boring, it takes an afternoon, and skipping it is the reason plenty of new sites publish for a year and get nothing.

1. Verify Google Search Console the week you launch, not the week you get worried. Prevents: having no idea what happened. Costs you: Search Console only shows data from the day you verify. Six months of history you never collected is six months of diagnosis you can't do.
2. Check you're not blocking Google. The staging noindex and the Disallow: / in robots.txt are meant to be removed at launch. Prevents: the single most catastrophic technical error there is. Costs you: literally all of your organic traffic, silently, until somebody happens to notice.
3. Fix indexing before you publish more. If Search Console shows 40 of your 100 pages indexed, adding 30 more pages just adds them at the same failure rate. Prevents: paying for content that can't appear in search. See why pages don't get indexed for the usual causes.
4. Make sure every page is reachable by clicking, not just by sitemap. Prevents: orphan pages. Costs you: slower discovery, and a page that receives no internal weight from the rest of the site, so it competes with one hand behind its back.

Tips 5–8: one page, one intent

Most content problems on small sites are really allocation problems. The writing is fine. It's just spread across pages that are quietly competing with each other, or aimed at a question the searcher wasn't asking.

1. One page per intent, permanently. If two URLs both want to rank for 'seo pricing', you haven't doubled your chances — you've split them. Prevents: keyword cannibalization. Costs you: Google picks one of the two, usually not the one you'd have chosen, and it ranks lower than a single merged page would have.
2. Match the format the results page is already showing. Search your query and look at the top five. If they're all comparison tables and you wrote an essay, you've answered a different question. Prevents: a well-written page that can reach position 15 and never move. Read search intent if that distinction is new.
3. Don't chase queries you'd never sell against. 'What is marketing' has enormous volume and no buyers. Prevents: traffic that converts at zero. Costs you: content budget, plus a topical signal that tells Google you're about something other than what you sell.
4. Answer the question before you tell the story. Put the direct answer in the first sixty words, then explain. Prevents: losing the snippet and the AI citation to whoever answered faster. Costs you: the readers who skim, which is most of them.

Tips 9–11: write the title for the results page, not the CMS

Your title tag competes with nine others in a list. It's read in about a second, next to rivals who have also had a go at it. Most titles lose that fight before the content is ever seen.

1. Write it for the searcher, not the admin panel. 'Services | Company Name' is how a CMS labels a page, not how a person decides. Prevents: ranking well and getting clicked badly. Costs you: clicks on positions you already earned, which is the most annoying way to lose traffic.
2. Unique title and description on every URL. Prevents: template-generated duplicates across 200 pages. Costs you: control — when your titles are unhelpful, Google rewrites them from your page content, and you no longer choose the one line searchers read.
3. Use the words people actually type, in the title and the H1. A clever headline that shares no vocabulary with the query makes Google work out the match. Sometimes it does. Prevents: throwing away relevance you'd already earned for the sake of a pun.

Tips 12–15: fewer pages, better, and stop counting volume

The instinct to publish weekly comes from social media, where the feed forgets you. Search doesn't forget you. A page from two years ago can still be your best-performing URL, which changes the maths completely.

1. Publish fewer, better pages. Four pages a month you'd happily send to a customer beat twelve you'd be embarrassed by. Prevents: a growing pile of thin content that dilutes what your site appears to be about. Costs you: the content budget, and then the cleanup budget later.
2. Update before you publish. Before writing a new post, check whether you already covered the topic. Prevents: a second page competing with a page that was starting to work. Costs you: the ranking the original page had.
3. Check intent before you check volume. A 20,000-a-month keyword whose results page is entirely Wikipedia, news and government sites is not an opportunity — it's a wall with a number painted on it. Prevents: the most reliably wasted money in beginner SEO.
4. Give every new page at least one internal link from an existing page on the day you publish it, from body copy, with anchor text that describes it. Prevents: publishing into a void. Costs you: weeks of discovery time and whatever authority the linking page could have passed.

Five tips you'll hear everywhere that don't hold up

These aren't fringe opinions. They're in most beginner guides, most YouTube explainers, and quite a few agency decks. They're still wrong, or at least badly out of date.