NOTE

The SEO advice I stopped following

Some of the most repeated SEO advice is not wrong so much as stale: it was a reasonable proxy once, then it became a rule, then it became the goal. This is a short list of the advice I stopped following, and what I do instead.

By Maharshi KushwahaUpdated 3 min read

Most bad SEO advice started as a decent observation. Someone noticed that pages which ranked well tended to be long, or that sites with many links tended to rank, and the observation hardened into an instruction. The instruction outlived the reason for it. Here are the ones I dropped.

Word-count targets

"Aim for 2,000 words" is the advice I hear most, and I stopped giving it years ago. The right length is whatever it takes to resolve the query completely and no more. A page answering "what is a canonical tag" should be short, precise and done. A page on migrating a site between CMSs should be long because the task is long. When I set a word count in a brief, writers pad. When I set the questions the page must answer, they write what is needed. I still write content briefs, but the brief specifies coverage, not length.

Publishing cadence as a KPI

"Publish three posts a week" turns a content team into a factory with no customer. Cadence is only useful if every piece has a reason to exist in the site's topic map. I would rather see one page a month that fills a real gap in a cluster than twelve that compete with each other for the same query and dilute the whole section. Volume is an input. It is not a strategy.

Keyword density

I have not counted keyword occurrences on a page in a very long time, and I do not think anyone should. Search engines have understood synonyms, entities and context for years. What matters is whether the page clearly addresses the topic and the intent behind the query, in language a person would actually use. Repeating the target phrase every hundred words makes the page worse for readers and does nothing for rankings.

Chasing domain authority

Third-party authority scores are a vendor's estimate, not a ranking signal. I have watched teams turn down a relevant link from a small, respected industry site because its score was low, and chase links from large irrelevant sites because the number was high. Relevance, placement and whether a human would plausibly click the link matter more than a metric Google does not use.

Exact-match anchor text everywhere

Forcing the target keyword into every anchor, internal and external, is one of the easier ways to make a link profile look manufactured. Internally, I want anchors that describe the destination naturally and vary across the site. Externally, I want links I did not control the anchor of, because those are the ones that actually indicate something.

"Content is king" without architecture

This is the one I feel most strongly about. Good content on a site with poor architecture is like a well-stocked warehouse with no shelving: the goods are there, nobody can find them. I have seen excellent articles sit unindexed because they were four clicks from anything, with no internal links pointing in. Content earns its value only when the structure around it lets crawlers and readers reach it, understand what it belongs to, and move on to the next relevant thing.

A rule is a proxy for a reason. When you forget the reason, you start optimising the proxy.

What replaced all of it

One question, asked before any piece of work: what does this change for a person searching, and for the crawler that has to find and understand the page? If the answer is "nothing, but the number goes up", I skip it. If you want the fundamentals that actually hold up over time, I have written them out in SEO basics, with the mechanisms rather than the rules.

Work with me

Need help applying this?

Reading is the easy part. If you want this implemented on your site, tell me what you're working on.

Work With Me Follow the SEO Roadmap
Email — [email protected]
Based in — India · working remotely
Response — usually within 2 working days