SEO Myths That Waste Your Time

Most bad SEO advice was true once. That is exactly why it survives: somebody watched it work in 2011, wrote it down, and it has been quietly costing people afternoons ever since. Here are twelve of the most expensive, what is actually true, and what to do instead.

Why these particular myths refuse to die

Three reasons, and recognising them will help you spot the next one on your own.

They were once accurate. They are easy to measure, which makes them satisfying to report on. And in several cases somebody sells a product whose value depends on you believing them.

A useful habit before adopting any tactic: ask what would have to be true for this to work, and whether Google has ever said anything resembling it.

The twelve

1. Keyword density has a target percentage

The belief. Your main keyword should appear in some magic share of the text, usually quoted as 1 to 2 percent, and tools will grade you against it.

What is true. There is no target and there never was a published one. Google matches meaning, not repetition, and has done since well before the phrase "semantic search" became a conference topic. Repeating a phrase to hit a ratio actively makes writing worse.

Instead. Put the phrase in the title and one heading, use it where a person would naturally use it, and then forget about it. If a tool tells you a well-written page is underoptimized at 0.4 percent, the tool is measuring the wrong thing.

2. The meta keywords tag still matters

The belief. Filling in a hidden list of keywords in the page source helps search engines understand the page.

What is true. Google said publicly in 2009 that it does not use the meta keywords tag for ranking web search results. That is not a rumour or an inference; it is an announcement, and it is older than most SEO agencies.

Instead. Spend the effort on the title tag and the meta description, which do real work. The title influences ranking and both influence whether anybody clicks.

3. Long articles rank better, so aim for 2,000 words

The belief. There is an ideal word count, and longer pages beat shorter ones.

What is true. Word count is not a ranking factor, and Google has said as much repeatedly. The correlation people cite is real but backwards: thorough answers to complicated questions tend to be long, so long pages tend to rank. Length is a symptom of coverage.

Instead. Open the top three results, work out what they collectively cover, and cover it better. Take a tile shop writing about underfloor heating under porcelain: if the top results answer nine questions and yours answers eleven with better photographs, the word count sorts itself out.

4. Publishing more often raises your rankings

The belief. Google rewards sites that publish frequently, so posting three times a week beats posting once.

What is true. Frequency is not a ranking factor. What frequency actually does is buy you more chances: more published pages means more queries you could rank for, which is a real and worthwhile effect that people then misattribute to a schedule. Freshness matters for queries where it matters, such as news, prices, and anything with a year in it, and not at all for a page about how underfloor heating works.

Instead. Publish at the rate you can sustain while still reviewing each piece properly. One good article a week for a year beats forty in March and silence until October, mostly for reasons that have nothing to do with the algorithm.

Publishing Cadence: How Often Should You Actually Publish?→

5. There is a duplicate content penalty

The belief. If the same text appears in two places, Google punishes your site.

What is true. There is no duplicate content penalty in the sense people mean, and Google has been saying so for well over a decade. What actually happens is duller: when Google finds the same information on several URLs it picks one to show and filters the rest out of the results. Nothing is punished; one version simply wins and the others become invisible.

Instead. Decide which URL you want to be the canonical one and say so with a canonical tag. The genuine risks nearby are different problems with different names: scraped content is a spam policy, and two of your own pages targeting one query is cannibalization.

Keyword Cannibalization: When Your Own Pages Compete→

6. Bounce rate and time on page are ranking factors

The belief. Google watches your analytics, and a high bounce rate drags you down.

What is true. Google has consistently said it does not use Google Analytics data for ranking, and bounce rate as your analytics defines it is not an input. The honest nuance, which most myth lists skip: documents and testimony from the US antitrust proceedings made clear that aggregated interaction data from search results themselves does feed the ranking systems. Those are two different things. What people do on a results page is not the same as what your analytics measures on your site.

Instead. Stop optimizing bounce rate as an SEO metric. A single-page visit where somebody got their answer in eleven seconds is a success, and treating it as a failure has produced a decade of pages that hide the answer to keep you scrolling.

7. You need to submit your site to search engines

The belief. New sites must be submitted, ideally to hundreds of engines and directories, or nobody will find them.

What is true. Google finds pages by following links and reading sitemaps. Submission services selling placement in 500 engines are selling access to search engines that either do not exist or do not matter. There is one legitimate version of this: verify your site in Search Console, submit a sitemap once, and use URL inspection to request indexing for a page you genuinely need crawled now.

Instead. Sitemap in Search Console, internal links from pages that already get crawled, done. If a new page is not getting indexed, the cause is almost always quality or orphaning, not a missing submission.

Google Search Console→

8. Sprinkle in LSI keywords

The belief. Google looks for "latent semantic indexing" keywords, so you should scatter synonyms and related terms through the text.

What is true. Latent semantic indexing is a real information-retrieval technique from the 1980s, designed for small static document collections, and it does not scale to the web. Google does not use it, and Google's own search advocates have said flatly that there is no such thing as LSI keywords. The term is an SEO invention that borrowed a real acronym to sound authoritative.

Instead. The underlying instinct is not wrong, just badly named. Cover the subject properly and the related terms appear because you cannot write about a veterinary clinic's dental service without mentioning anaesthesia, scale and polish, and extractions. Writing about the whole topic is the technique; a list of synonyms to sprinkle is not.

9. More pages always beats fewer

The belief. Every additional page is another lottery ticket, so publish everything.

What is true. Beyond a certain point, extra pages cost you. Thin pages compete with each other, dilute internal linking, and give a site-quality assessment more weak examples to average over. Volume produced primarily to rank rather than to help anybody is the shape Google's scaled content abuse policy is written to catch.

Instead. Ask of each planned page: is there a real question here that somebody asks, and can we answer it better than what currently ranks. A language school with 40 genuinely useful pages about exam preparation will beat one with 400 near-identical city pages, and it is not close.

10. More backlinks is the goal

The belief. Rankings are a function of link count, so buy or build as many as possible.

What is true. Link quality and relevance carry the signal, and manipulated links are now mostly neutralized rather than punished, which means a bought link often does nothing in either direction. For long-tail and narrow B2B queries, links are frequently not the constraint at all.

Instead. Fix internal linking first, since you control every one of those. Then earn the handful of links that come from doing something worth mentioning. There is a whole page on this, including where links genuinely still decide the outcome.

Do Backlinks Still Matter in 2026?→

11. Get your Domain Authority up

The belief. DA is your site's score, and raising it raises your rankings.

What is true. Domain Authority is a Moz product, Domain Rating is an Ahrefs product, Authority Score is a Semrush product. Google made none of them and sees none of them. They are useful for comparing SERPs at a glance and useless as targets, because the fastest way to move one is exactly the link-buying that does nothing for rankings.

Instead. Track how many of your pages sit in positions 11 to 30, and how many distinct queries you get impressions for. Both move earlier than traffic and neither can be bought.

Domain Authority Is Not a Google Metric→

12. Core Web Vitals are a major ranking lever

The belief. Getting your speed scores green is one of the big wins available to a content site.

What is true. Core Web Vitals are a confirmed page-experience signal and they are genuinely small next to relevance and content quality. They behave like a tiebreaker between comparable pages rather than a lever. The metrics themselves also move: INP replaced FID in March 2024, so guides written before then are measuring something that no longer exists.

Instead. Fix the things that are bad for humans anyway. Compress your images, stop loading five tracking scripts, and give images explicit dimensions so the layout stops jumping. Then leave it alone.

Two habits that catch the next one

New myths appear faster than lists like this get updated, so the useful thing is not the list.

• Ask where the number came from. Any specific figure about Google's algorithm, a weighting, a threshold, an ideal count, is somebody's estimate at best. Google publishes guidance and policies, never weights.
• Ask who profits from you believing it. Speed myths are told by speed tools, link myths by link sellers, word-count myths by content mills paid per word. It is not always cynical, but it is always worth noticing.

What SEO cannot fix

The last thing worth saying is the one nobody in this industry gets paid to say.

SEO is a distribution channel. It decides how many of the right people arrive. It has nothing to say about what happens next, and there are three situations where no amount of it will help.

• A bad offer. If people arrive, read, understand, and still do not want it at that price, more of them arriving changes nothing except your hosting bill. A wedding photographer whose portfolio is weak does not have a traffic problem.
• No demand. If nobody is searching for the problem you solve, there is no traffic to capture, and the honest strategy is a different channel entirely. Search is for demand that already exists; it is a poor tool for creating a category.
• A site people cannot buy from. Broken forms, a phone number nobody answers, three-week quote turnarounds, checkout that fails on mobile. Traffic makes these more expensive, not less.

There is a fourth, quieter one. If you are not the best available answer to the question and you have no plan to become it, ranking first is a temporary accident.

Every myth in this page is a way of avoiding that sentence. The percentages, the word counts, the scores out of 100, all of them offer a version of SEO where you can win without being better at the thing.

The work is being better at the thing. The rest is distribution.

How Long Until SEO Works? A Realistic Timeline→ SEO Terms, Explained→ What a Keyword Really Is→