LSI (Latent Semantic Indexing) keywords are keywords that are related to a main or central keyword in a given context. For example, let us say that our main keyword is ‘swimming’. A set of LSI keywords for ‘swimming’ may include ‘wade’, ‘dive’, ‘float’, ‘ocean’, ‘deep’’, and ‘backstroke’.
In the example above, some of the LSI keywords may be synonyms for the main keyword (e.g. wade), but this is not necessary. The main idea is that all of these keywords are likely to be found in an article about ‘swimming’, and are related to it.
As its name suggests, LSI is an indexing technology in natural language processing, and was invented by Susan Dumais at Bell Labs in the late 1980s, and predates the invention of the Internet.
One of the ongoing debates in recent times has been about whether Google uses LSI Keywords. While some members of the SEO community believe that LSI Keywords are integral to content marketing and on-page SEO strategies, the growing consensus is different.
Here’s what Google says about it:
Based on the above tweet from Google’s John Mueller, it is an SEO myth that Google uses LSI Keywords. This means that Google's algorithms do not use the underlying mathematics of Latent Semantic Indexing to understand content.
However, Google, or any search engine, for that matter, needs to understand the semantics around content that is on the web in order to match it with a user’s search intent. Therefore, if you want to apply search engine optimization to your own content, or you are conducting keyword research for your ads, it remains extremely important to identify the keywords that are semantically related to your primary keyword.
There are a few different tools that can be used to identify keywords that are semantically related to a primary keyword.
Here are a few:
An easy way to obtain keyword ideas is to use Google’s AutoComplete feature. As an example, typing in the keyword football into the Google search bar, returns a bunch of related words such as ‘today’, ‘schedule’, ‘games’, ‘playoffs’ and so on.
This section of the Google SERP shows a list of questions that are related to the primary search. For example, a search for ‘how to make espresso’ shows the following related questions
These questions give us some indications about related keywords, such as ‘regular coffee’, ‘machine’, ‘step’ and ‘beginners’.
Another way to obtain keywords that are related to the primary keyword is by using Google’s own Keyword Planner. By entering the primary keyword, you can find related keywords. From the example below, we can see that ‘without machine’, ‘instant coffee’ and ‘cortado’ are related keywords.
Although the free approaches above are useful in finding semantically related terms, they do not provide their relative importance. This means that simply focusing on some of the keywords suggested by these tools will not help your content rank better or make your ad campaigns more successful.
This is where a tool like TF-IDF Tool can add immense value to your SEO strategy. This is a highly optimized version of the TFIDF algorithm that simplifies the process of obtaining a list of highly relevant keywords that you should focus on, already ranked for you by importance. Moreover, TF-IDF Tool eliminates all the noise of keywords that may be related to your primary keyword, but are not contextually relevant.
Here is how it works:
1. From Google search results, TF-IDF Tool takes the top 20 websites that rank for the primary keyword and extracts the statistically important terms found in their content
2. It generates TF-IDF scores for each term and sorts them
3. It summarizes the analysis with a clear list of terms that are closely associated with your target keyword.
Best of all, it does this within a couple of minutes, saving you a ton of time, and gives you the results in a nicely packaged PDF or CSV file. You can immediately use this list of terms either in determining which keywords to target, or in identifying opportunities for content optimization and creation.
Here is a quick comparison between TF-IDF Tool and another tool in the space:
Looking at the results side-by-side, here are some neat advantages of TF-IDF Tool:
1. The results from TF-IDF Tool include scores for each keyword, so you know what to focus on
2. The results are based on a real-time analysis of top ranking sites, this means that each keyword in the list is guaranteed to be related to the primary keyword.
3. TF-IDF Tool suggests keywords that are relevant in time (‘tokyo 2020’, ‘paris 2024’), and also semantically associated (‘adam peaty’ refers to the well-known swimmer who specializes in the breast stroke) with the keyword ‘swimming’.
4. Finally, the suggestions include long-tail keywords, which can significantly help with your content ranking for specific searches where the user is closer to the point-of-purchase.