LSI Keyword Tactics for Advanced SEO Experts


Enhance Keyword Research with LSI

Google and other search engines are continuously rolling out updates to improve themselves. Their goal is to deliver the best user experience to their customer, which means to develop an algorithm that will return the most relevant results for the queries being searched.

Panda… Penguin… This zoolandish rollout of algorithmic changes incorporated many different methods for returning higher quality content to the user, while weeding out the low quality articles methodically written to rank high in search engines. Gone are the days when you could rank for a particular term by stuffing keywords into your webpage. Google has come to understand that natural writing typically does not consist of the redundant word use of a keyword phrase, and this has been a huge game-changer to what elements influence an article to rank high in the SERPs.

Welcome to the new era of SEO Writing. This article focuses on how to utilize latent semantic indexing  to your advantage in SEO, explained though a theoretical dichotomy I’ve adapted to effectively implement it in a SEO Strategy. Although intricate data is provided about LSI in Google’s patent: Information retrieval and text mining using distributed latent semantic indexing, it seems the closest consumer-friendly mentioning  that Google has on this topic refers to their Knowledge Graph project. You can read more about the Knowledge Graph in their article: Introducing the Knowledge Graph: things, not strings or watch their video about it.

What is Latent Semantic Indexing (LSI)?

LSI is the part of Google’s algorithm that will fold words into a matrix for analyses, for the purpose of identifying semantically related terms. Search engines apply this process through linking related words together by analyzing how certain keywords on webpages appear in combination with other related keywords, then correlating them to a particular theme or concept. LSI is the reason why search engines are smart enough to see that that Tiger Woods plays golf; and that Saddam Hussein is somehow closely related to Iraq and the Gulf War. LSI also allows search engines to better decipher webpages to provide more relevant search results with higher quality content, free from keyword stuffing and spam.

(of a quality or state) existing but not yet developed or manifest; hidden; concealed. Also: dormant, hidden, concealed, underlying, invisible, or unseen.

relating to meaning in language or logic.

Something that serves to guide, point out, or otherwise facilitate reference.

A great example of this, taken from an NCSU research paper on Semantic Search, illustrates complications a search algorithm may have searching the query Mississippi River Bank. From the large archive of indexed webpages, an algorithm may have difficulty determining the most relevant result to return. Algorithmically, this string could be interpreted as a riverbank in Mississippi, semantically relating to fishing and recreation; or a financial institution in Mississippi, semantically relating to mortgage loans and banking. The latent semantic indexing may look something like this:

Mississippi Search LSI Example

Basic example of Latent Semantic Indexing

In application to SEO, implementing LSI keywords will enhance content to rank better then from high keyword density alone. It also decreases the chance of not getting ranked from keyword stuffing, as well. The idea behind this paradigm is that natural writing does not redundantly use keyword phrases referring to it’s articles topic (keyword stuffing), but instead, uses a variety of synonymous and relevant keywords relating to its topic. SEO experts follow the idea of LSI by implementing synonymous and relevant terms of their main keyword. This SEO tactic adheres to the LSI concept that webpages are composed of certain keywords that commonly appear with other related keywords. Therefore, if writers want to rank for a particular keyword phrase then they should make sure that their content contains other related LSI terms.

How Do You Find LSI Keywords?

I have made the distinction that LSI keywords can be categorized into two buckets; Synonymous LSI (sLSI) keywords and Relating LSI (rLSI) keywords. The Synonymous LSI keyword bucket should be your main focus as an SEO Expert. However, it’s important you understand the difference between these two groups.

Suppose you write an article targeting the keyword Apple. The Relating LSI Keywords that Google identifies will help to determine whether your article is about the brand or the fruit. If keywords such as iPhone, Mac OS, or Phone Accessories are found, Google will semantically index it with the brand Apple, and rank it accordingly. Alternatively, if keywords such as Pie, Recipes, or Juice are found, Google will know it’s referring to the fruit.

Historically, low quality SEO Articles tend to be narrowly written and keyword stuffed in order to quickly produce and rank content. Articles containing a variety of rLSI keywords can be used as an indicator by search engines that it was naturally written. Inserting rLSI keywords in your content to complement its targeted keyword phrase is a great SEO tactic, especially if you are trying to rank for competitive keyword terms.

Method of Discovery:

There are a few services that offer amazing research tools to assist in finding rLSI keywords. Although each works a bit differently, their methodologies are alike: First, Provide the keyword you want analyzed. They search that term in Google, and begin crawling the SERP list, scrapping the text of every article that ranks high. This text is then combed through, removing common and irrelevant words such as and, or, this, the, etc. It then returns a keyword list of the most commonly used terms found in the collection of articles it scraped. This list serves as an amazing reference to possible rLSI keywords, and can be used to complement your targeted keyword phrase and diversify your content.

Two services that do this are InboundWriter and Inbound Writer has an amazing algorithm that will return a high quality list. However, it is a paid service, and also primarily functions to help SEO writing. is free, but returns a list that will require some manual sorting. Many others can be found on the internet as well, including Wordstream (paid) and the Google Sets Method (amazing).

Another method is to do a search in Google yourself. At the bottom of your search page Google provides suggested searches. These are clues to what Google is relating to that keyword. As you can see, the related searches for my phrase best face cream introduced a bunch of new terms associated with my keyword phrase, such as anti-aging, wrinkles, and dry skin.

Example of Google Suggest relating to LSI for SEO

Synonymous LSI Keywords (sLSI)

This group dissects latent semantic indexing at a micro-level. sLSI terms are a lot more difficult to target and locate. However, if you can find them, it’s the way to power-charge your articles for good rankings. Google will make direct synonymous correlations between terms, in order to maximize on returning a better set of relevant results. sLSI impacts search results at a keyword-level. One can observe that Google places the same weight in sLSI keywords as to its synonymous counterpart. You can read more technical information about this concept in Google’s article: Helping computers understand language. I’ve noticed that using sLSI keywords to complement its archetype potentially enhances its ranking ability. The same applies for contextual link building as well.

Method of Discovery:

Finding sLSI keywords is simple, but requires some manual effort. To being, you must first notice that Google always returns its search results list with keyword terms bolded throughout the page. You will also notice that synonymous terms are bolded as well; this is what indicates its an sLSI term.

Lets take a look at a basic example. When searching 2 examples of lsi, you will notice that Google made an association between the phrases LSI and latent semantic indexing, as well as 2 and two – as all are bolded in the results. Even more interestingly, the article that ranked highest contained the searched keyword phrases and its sLSI counterparts.

Example of LSI Keywords

As you can see, the way to identify your sLSI keywords are simply by searching your keyword term in Google, and looking for the word variations that have been bolded.

A Deeper Look into sLSI

Although a basic concept, sLSI keywords are the power-players in SEO. In this next example, we look for the sLSI equivalents for the keywords Best Face Cream. You will find sLSI keywords for each individual word.

Best = Top | Face = Facial | Cream = Moisturizer


Keep in mind, this task isn’t always so easy. Many of times I have struggled to find the sLSI equivalent for my keyword. One example is the keyword Testosterone, which does not have any sLSI keyword associations. In situations like these, I use the best rLSI equivalents in my content, which in this case was Low-T, amongst a few other hormones.

Other Methods of Discovery:

Google Adwords can sometimes be an effective method for finding sLSI and rLSI possibilities. To do so, go to the Keyword Planner tool in Adwords. After entering your keyword phrase, use the negative search attribute on the side bar against the term you are searching.

LSI Adwords SEO

Complementing your keyword phrases with sLSI has a great potential to supercharge your articles. You can also implement sLSI keyword terms as a Panda/Penguin Safe way of link building.

How Do You Create SEO-Friendly Content With LSI?

For in depth instructions, please read my article on SEO Writing using LSI.

  1. Of course, you have to pinpoint the main keyword or main keywords you want to target (keyword Research).
  2. Then, look for rLSI and sLSI terms to that of your main keywords (instructions above).
    • rLSI will be used to diversify your articles and broaden it’s information.
    • sLSI will be used to enhance the keyword density of the term you are trying to rank for.
  3. Third, create content with a double purpose, one with the viewers in mind, the other with the search engines taken into consideration. So, how do you do that?
    • Scan through all the related terms (rLSI) and look for concepts or themes that are important and related to the main keyword. After which, you create content covering all of these themes and concepts. In that way, you will have the liberty to create useful content that makes use of LSI keywords.
    • Share keyword density of your keywords with your sLSI phrases. Use sLSI keywords slightly less than your main keyword phrase.
  4. And finally, continue to use SEO Friendly methods of content writing, such as including your keywords in your H1 and H2 tags, make sure your keywords are used at lease once above the fold, and implement your keywords in your meta-data (title tags) and semantic-markup as well.

Complementing your keywords with LSI provides great potential to supercharge your articles. Also consider sLSI keyword terms as a Panda/Penguin Safe way of link building; to diversify your anchor text. I believe current SEO strategy has shifted focus to providing high-quality search results to its users. If you couple praiseworthy content with a good LSI strategy, then your content will really get a search engine boost.

Request Consulting on This Topic

NOTE: Jourdan Rombough coined the terms rLSI and sLSI, which have not been acknowledged by Google or other industry experts. They are terms created to better explain his methodology of LSI keyword research and its implementation into SEO writing.

About Author

Jourdan Rombough is a SEO Expert specializing in Website Visibility & Internet Strategy. He is an SEO Manager at MERKLE | IMPAQT, assisting Fortune 500 Companies with their visibility strategies. More...


  1. Nice write up Jourdan. Totally agree with you, if you want to survive from Panda and Penguin , LSI can be your savior. Now Keyword Density concept has gone. If you want to rank you have to write natural, write for the user not for the crawler. If you implement LSI, your content will most likely look natural.

    One quick or you can say a funny question: What if i write an article on “APPLE”, i used word apple and both of its its related synonyms like eatable apple, red apple, apple tree and related synonyms like ipad, iphone etc. Google associate this article with which apple? An apple or Apple company? 😛

    • They work AMAZING for back-linking. And can be used as a penguin/panda friendly safe way to link build, diversifying anchor text redundancies in contextual links. I have personally seen great results!

  2. Unfortunately kwmap-dot-net no longer goes to the keyword tool that it used to. It is now about, well, you can check it out if you want to. But if I were you, I wouldn’t be linking to it.

Leave A Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.