SEO Myths: Which Are False, Which Are Expired

Updated September 2026 · Written and maintained by the Progression Agency strategy team

Almost no SEO myth was invented. Most were accurate descriptions of how search worked at some point and have outlived the system they described; a smaller number were never true; and a third group are entirely true principles that get turned into rules nobody should follow. Sorting a claim into one of those three buckets is more useful than arguing about it, and it usually takes about five minutes with a primary source.

Why myths about SEO are so persistent

Answer first: because most of them were true once. Search changes continuously and articles do not, so advice keeps circulating long after the system it described has gone. Myths about SEO spread faster than corrections because they are simpler than the truth.

Simple claims travel further than accurate ones

A rule with a number in it is easy to repeat, easy to check against and easy to sell. The accurate version is usually conditional, which makes it harder to remember and far less quotable.

The three kinds of SEO myth

Answer first: false claims, expired truths and misapplied truths. They need different responses, which is why lumping them together as myths makes the conversation worse rather than better.

Expired truths are the largest category

Search has changed continuously, and advice written honestly in one period keeps circulating long afterwards. This is why the publication date of an SEO article is more informative than the reputation of whoever wrote it.

Misapplied truths do the most damage

Nobody wastes much money on meta keywords. A great deal gets wasted on publishing volume, arbitrary word counts and optimizing pages that were never going to rank, all of which descend from principles that are perfectly correct.

Myths that are simply false

Answer first: keyword density targets, meta keywords, a duplicate content penalty, mandatory search engine submission, domain age as a factor, and Google ranking you on Analytics data. Each is contradicted by published documentation.

There is no duplicate content penalty

Where the same content appears in several places, search engines pick one version to show and filter the others. That is selection, not punishment, and the practical consequence is that the wrong version may be the one shown — which is a canonicalisation problem rather than a penalty. Google’s guidance on consolidating duplicate URLs describes the mechanism.

Analytics data is not a ranking input

The belief persists because it feels plausible that a search engine would use the behavioral data it collects. Treating it as true leads people to manipulate bounce rate and time on page, which changes a number in a report and nothing in the results.

Domain age correlates with rankings and does not cause them

Old domains tend to rank because they have had longer to accumulate links, content and reputation. Buying an old domain does not transfer that, and expired domain abuse is explicitly listed in the spam policies.

What the search engines actually publish as spam

Answer first: read the policy rather than an article about it. Google’s spam policies name specific behaviors including scaled content abuse, link spam, cloaking, doorway pages, site reputation abuse, expired domain abuse, hidden text, keyword stuffing, sneaky redirects, thin affiliate pages, user-generated spam and misleading functionality.

The authoritative list is at Google Search Central — spam policies, and the broader requirements sit in Google Search Essentials. Both are short, free and more reliable than any summary, including this one.

Using AI to help write is not on the list

What is on the list is scaled content abuse — producing many pages primarily to manipulate rankings rather than to help people — regardless of whether a human or a machine wrote them. The distinction is purpose and quality, not the tool.

A slow page is not spam

It is a usability problem that can cost you visitors and, at the margin, rankings. Conflating performance issues with policy violations produces panic about the wrong things and complacency about the right ones.

Truths that get misapplied

Answer first: content quality, links, speed, mobile usability, intent matching, internal linking and structured data are all real. Publishing frequency, arbitrary length and blanket optimization are the proxies people substitute for them.

Longer content does not rank better

Length correlates with ranking because thorough answers to complex questions tend to be long. Padding a page that answers a simple question adds words and removes clarity, and the correlation does not survive being used as an instruction.

Publishing frequency alone does nothing

Sites that publish often tend to rank well because publishing often is what an organization investing in content looks like. A schedule filled with thin posts produces the appearance without the cause, and increasingly attracts the scaled-content problem instead.

Structured data affects eligibility, not position

Marking up a page can make it eligible for particular result features, which can materially change how it appears. It does not raise the underlying ranking, and adding schema to a page that does not deserve to rank changes nothing.

What actually moves rankings

Answer first: matching the search intent, being genuinely useful, earning links from real sites, being crawlable, and linking internally with meaningful anchors. Speed matters at the margin.

Intent matching is the largest lever and the least discussed

A page can be excellent and still lose to a worse page that answers the actual question behind the query. Looking at what currently ranks and asking what those pages have in common tells you the intent faster than any tool.

Fix what already ranks before writing anything new

Pages already receiving impressions are the cheapest improvement available, because the audience is already arriving and the page is already eligible. New content is more visible in a report, which is why this obvious step is so often skipped.

How to test any SEO claim yourself

Answer first: find the primary source, check the date, ask what the claim predicts, ask what the mechanism would be, ask who benefits from believing it, and then test it on one page rather than the whole site.

Ask what the mechanism would be

For a search engine to use a signal, it has to be able to observe it. That single question eliminates a large share of bad advice immediately, because many claimed factors describe things the engine has no access to.

A claim that predicts nothing cannot be tested

Advice phrased so that any outcome confirms it is not a claim about the world. If rankings going up and rankings going down would both be taken as evidence, there is nothing there to evaluate.

Who benefits from each myth

Answer first: some myths survive because somebody sells the remedy. That is not a conspiracy, it is an incentive, and noticing it is a fast way to sort advice.

Scores are not rankings

Tools that produce an optimization score out of a hundred are measuring adherence to their own checklist. A page scoring ninety can rank below one scoring forty, because the checklist is a proxy for factors nobody outside the search engines can observe directly.

A long audit is not the same as a valuable one

Automated audits list every deviation from a default, which produces hundreds of items of wildly varying importance. The useful version identifies the three things holding the site back, which requires judgment rather than a crawler.

When advice stopped being true

Answer first: meta keywords, exact-match domains, keyword density, article directories, guest-post networks and separate mobile sites all had a period of genuine effectiveness that has passed. Recognizing the pattern is more useful than memorizing the dates.

The pattern repeats

A tactic works, it gets systematized, the systematization gets detected, and the tactic becomes at best useless and at worst a liability. Anything currently described as an easy win is somewhere on that curve, and the question worth asking is where.

Myths specific to local search

Answer first: geotagging photographs does nothing, posting daily does very little, adding every available category dilutes rather than helps, and keyword-stuffing a business name works and is the most common cause of suspension.

• Geotagged images do not influence local ranking
• Posts are real and among the smallest levers
• Extra categories dilute the primary signal
• Proximity dominates and cannot be bought
• A keyword-stuffed name is reportable by any competitor
• Duplicate profiles are consolidated or suspended, not additive

Name stuffing is the exception that proves the rule

It genuinely does affect visibility, which is why it persists. It is also against policy, trivially reportable, and the asset at risk is the profile producing your calls. Most local myths cost you nothing to believe; this one is different because it works until it does not.

Myths about links

Answer first: links still matter, buying them is a policy violation regardless of relevance, disavowing is not routine maintenance, and the number of links matters far less than where they come from.

Relevance does not make a paid link acceptable

Link schemes are named in the spam policies, and a paid link on a relevant site is still a paid link. The rationalization that relevance makes it safe is the most common form the myth takes.

Disavow is a rare tool, not a monthly task

It exists for specific situations, typically involving a manual action or a genuine attack. Routine disavowing of ordinary low-quality links that were never solicited is unnecessary and occasionally harmful, and it is frequently sold as maintenance.

Myths about technical SEO

Answer first: technical work makes a site eligible to rank and rarely causes it to rank. It is necessary and it is not sufficient, and the industry oscillates between over-weighting and ignoring it.

Sitemaps do not guarantee indexing

They help discovery, particularly on large or new sites. A page in a sitemap that the engine judges not worth indexing stays unindexed, and repeatedly resubmitting it does not change that assessment.

Myths about how long SEO takes

Answer first: three to six months for new content in a competitive space, weeks for improvements to pages already ranking, and days for technical fixes that were blocking indexing. Anyone promising faster on new content in a competitive area is describing something else.

The timeline depends on where you are starting

A site with existing authority publishing into a topic it already covers moves far faster than a new site entering a competitive area. Quoting a single universal timeline is one of the more harmless myths and it still produces badly set expectations.

Myths about content and AI

Answer first: the question is never who wrote a page but whether it was produced to help somebody or to occupy a search result. Myths about SEO and AI mostly collapse once that distinction is applied.

The line is scale and purpose

A few genuinely useful pages assisted by a model are not the problem. Hundreds of pages produced primarily to capture search demand are, and that was equally true when the same thing was done by cheap human writers a decade ago.

Myths about analytics and reporting

Answer first: rankings are not a business outcome, position tracking is directional at best, and traffic that does not convert is a cost rather than a result.

Rank tracking is directional, not precise

Results are personalized and localized, so the position a tool reports is one sample of many possible results. It is useful for spotting trends and misleading when treated as a precise measurement of where you sit.

Impressions and traffic are not the objective

A page can gain a great deal of traffic from a query that will never produce a customer. Reporting that as success is one of the more comfortable myths in the industry, because it is easy to produce and hard to argue with in a meeting.

Myths about competitors

Answer first: you cannot see a competitor’s traffic, only an estimate of it, and copying their tactics assumes their economics match yours. Both are treated as facts far more often than they should be.

Domain authority is not a Google metric

It is a third-party model built to approximate something Google does not publish. It is a reasonable rough filter and it is not a score anybody at a search engine sees or optimizes against, which makes treating it as a target a category error.

Frequently repeated claims, briefly settled

Answer first: a short list of things people ask about constantly, with the honest one-line answer each deserves.

Frequently asked questions about SEO myths

Search, measurement and marketing talks from their publishers

Publicly available sessions on search, analytics and marketing measurement. None of these are ours; each is credited to its channel by name and upload date, every identifier was checked live before publication, and each tile loads its player only when clicked.

Why these myths persist

Most SEO myths are former truths. Meta keywords worked, keyword density approximated relevance, and directory submission was how pages were found. Each stopped being true years apart, and the advice outlived the mechanism because the people repeating it never had to test it.

Frequently asked questions