It used to be, if you wanted to rank web pages, then you had to count your keywords. Webmasters used what they called keyword densities.
Suggested keyword density was 1% to 5%, which means for every 100 words of content on a page, the keyword you were optimizing for had to appear 1 to 5 times on that page. It was not a very precise way of search engine optimization, but it worked, primarily because the search engine algorithms weren’t very sophisticated at the time. Today, the search engines look at many more ranking factors, one of which is related words.
This is called Latent Semantic Indexing, or LSI.
In a word, LSI means looking for related words to understand what a page of content might be about. For instance, if you were writing a page about raising chickens, what words might you use that would be associated with the subject matter without actually saying “raising chickens?”
Think about synonyms for the word “chickens.” What comes to mind? Roosters, hens, pullets. I’m sure you can think of a few more.
Now, how about “raising?” “Farming” comes to mind.
Latent Semantic Indexing is a form of natural writing that allows content writers to write more naturally and get pages ranked for key phrases even if those pages may not appear on your web page. A more effective natural writing technique might be to spread these semantic keywords around. Instead of relying on a single keyword mentioned 1 to 5 times for every 100 words of content, you might use 4 or 5 related words fewer times each. The search engine then puts all of those words together to determine what your page is about and ranks your page accordingly.
Welcome to the future of SEO.