Few concepts in SEO are as misunderstood as latent semantic indexing. This leads to a lot of misguided methodology from people keen to look like they know what they’re doing.
Part of the reason why it is so frequently mentioned by certain SEO ‘experts’ is because it sounds faintly scientific. People don’t understand the words, so they can explain it how they choose. In reality, it’s a fairly straightforward concept.
It is, quite simply, the indexing of material according to the meaning (semantic means ‘related to meaning’) that is present (latent) in the text.
That’s it. Indexing stuff according to meaning.
Here’s an example. The following words appear on a page:
• Search engine
• Docs
• Translate
• Analytics
• AdSense
From the meaning of those words, we can put together a complete picture and quite confidently state that this page is about Google. We don’t actually need the word ‘Google’ to appear to tell us that.
Latent semantic indexing revolves around the relationships between different words and the pictures those relationships build. Anyone who attempts to manipulate their web content to better exploit it for SEO is missing the point entirely, because the point of latent semantic indexing is this: it rewards plain, simple, on-topic writing.
Recent Comments