OK, so you heard the bell ringing but you not sure where the clock tower is. You may ask “what is ‘Latent Semantic Indexing’ and how do I implement this to strengthen my overall SEO strategy?
Latent Semantic Indexing or simply LSI forms the basis of the current Google algorithms and as such must be your underpinning SEO strategy. LSI is very much an overriding strategic foundation rather than an SEO technique. It should form the basis of everything you do in SEO.
So what is Latent Semantic Indexing?
In a nutshell, Latent Semantic Indexing falls in line with the concept of “siloing”. It basically suggests that you departmentalise the structure of your website. Within this structure, you heavily link related topic and equally important, not link unrelated topics. I.e., where you want to sell apples, talk about apples, apple varieties, apple diseases, apple cider, apple juice etc. Equally important, don’t talk about oranges or orange rind – if that makes any sense to you.
Another key characteristic of Latent Semantic Indexing is the fact that it focuses on the context in which the key word occurs and “rewards” this context often higher than the keyword itself.
For example, look at the following text optimised for the keyword “apples”:
“We grow the best Fuji apples within the established apple region of Huonville Tasmania…”
“We grow the best apples and oranges in the Riverina region in New South Wales…”
In the first text, the words “Fuji” add strength to the keyword “apples” because of the brand it represents. Equally, the geographical “Huonville” and “Tasmania” are associated with the keyword “apples”. Again, under the Latent Semantic Indexing principle, these words add value to the principal keyword “apples”.
In the latter example, the word “oranges” dilutes the value of the keyword “apples”. Equally, “Riverina” is associated with oranges and not apples so this word too may reduce the value of the keyword “apples”.
So when writing your content, try to put all your apples in one baskets and separate them from your oranges 🙂
Latent Semantic Indexing by Rene Nusse