Home Blog Latent Semantic Indexing (LSI) 101 for Enterprise Local SEO

You may hear the term latent semantic indexing (LSI), sometimes referred to as latent semantic analysis, thrown around in SEO circles. The question then arises, is LSI something you should be using to improve your online visibility? In this post, we take a look at what latent semantic indexing really is, why it matters, and how it affects your enterprise local SEO strategy. 

What is Latent Semantic Indexing?

For this, we go straight to the source: 

Susan Dumais is a technical fellow and the managing director of Microsoft Research New England, Microsoft Research New York City and Microsoft Research Montreal. Prior to joining Microsoft in 1997, while with Bellcore, she co-developed a statistical method for concept-based retrieval known as Latent Semantic Indexing.

In the overview of her research, she explains, “We assume that there is some underlying or ‘latent’ structure in the pattern of word usage across documents, and use statistical techniques to estimate this latent structure.”

In short, Latent Semantic Indexing helps search engines understand the relationships between words.

Or, as Clark Boyd succinctly explained for Search Engine Journal, “In essence, it finds the hidden (latent) relationships between words (semantics) in order to improve information understanding (indexing).”

Why Does LSI Matter in SEO?

Remember when a Google search results page was simply ten blue links and using a certain density of the right keyword in the right places could get you to the top of the pack?

Search has undergone an incredible evolution and maturation since those early days. Deepening the algorithm’s understanding of the relationships between words helps Google understand the relevance of any particular page on the web to the query at hand.

But does Google actually use Latent Semantic Indexing?

Not according to John Mueller, Webmaster Trends Analyst at Google:

Latent Semantic Indexing example

Latent Semantic Indexing matters in SEO simply because it’s a term that many have become accustomed to using to describe machines’ understanding of how words are related. But if Google isn’t actually using it, what does it matter?

If Not Latent Semantic Indexing, Then What?

It’s true, Google isn’t using LSI. It’s an older technology and while it almost certainly informed what’s been developed since then, Google doesn’t need LSI. In 2015, Google announced a major algorithmic update called RankBrain that uses machine learning and AI to better understand and process search queries.

Moz explains the purpose of RankBrain: “…it is believed that the query now goes through an interpretation model that can apply possible factors like the location of the searcher, personalization, and the words of the query to determine the searcher’s true intent.”

Keep in mind that Google isn’t the only search engine. Yes, it has over 73% of search engine market share worldwide (as of early 2020). But remember that Susan Dumais, creator of LSI, has been with Microsoft all these years and if Bing isn’t using LSI, her earlier work has almost certainly influenced the way information is indexed there.

Whether old-school LSI or more sophisticated AI-enabled algorithms are in play, one thing is certain: search engines want to understand not only the term but the intent behind it. Understanding that LSI and RankBrain exist gives us useful context as to what search engines are looking for as they decide which result to show searchers.

What Does This Mean for My Enterprise Local SEO Strategy?

You can help search engines understand that your Local Pages, listings and other content assets are the best result for local searchers by:

Following industry-recognized SEO best practices.

This entails optimizing local page titles, meta descriptions, H1 tags, image alt text, and page content with rich, descriptive content that appeals to search engines and readers alike. Provide a good user experience and link internally with your customer’s journey in mind. 

Using schema (structured data).

Schema markup helps search engines better understand the content of your site and can give you greater control over how various types of results appear in search. See Advanced Schema Strategies for Local SEO for more insights and schema best practices.

Localizing the content and structure of your pages.

Use localized, unbranded keyword terms throughout your metadata and page content to help Google understand your exact place in the local search ecosystem.

Incorporating long-tail keywords into your optimization strategy.

According to Ahref’s, 92% of the 1.9 billion search terms in its US database are long-tail keywords that get fewer than 10 searches per month. Get our Go-to Guide for Long-Tail Keywords Strategy For Local Success to learn how to use them effectively.

Improving the mobile-friendliness of your site.

Google puts a great deal of weight on mobile-friendliness and its no wonder. Their own research has found that 29% of smartphone users will immediately leave a site if it’s too clunky or slow, and 67% will find another answer if it takes too many steps on mobile to complete their purchase or get the information they need. Download our recent whitepaper, Mobile Local Playbook: What You Need to Succeed in 2020, to learn more.

Optimizing for different types of search, including voice search.

Google is making it easier for consumers to listen to longer sections of text via their voice assistants. How you format and craft content can also impact its friendliness for voice search

Displaying online reviews on your local pages.

Online reviews are valuable user-generated content that contains keywords and sentiment Google uses in its evaluation of your page. See 5 Key Local Reviews Opportunities for Enterprise Brands for more on making the most of your online reviews strategy.

Learn more about local SEO strategy and tactics for enterprise brands: