LSI. Sounds like a new term? Keep on rolling. Latent Semantic Indexing (LSI) is a part of the Google algorithm. Knowing about it and implementing it as a part of your SEO strategy is crucial. This post presents you the concept in an easily understandable way.
Definition, "LSI keywords (also known as semantic keywords) are the related keywords to the main keyword. Many of the search engines like Google use it to decide the link between different entities of the web content." It helps Google get the meaning of the written content.
It also deals with search relevancy. Including LSI keywords with main keywords is SEO friendly. In earlier days, search engines used exact search query to get results. With this, search engines were not giving accurate results. Now, search engines analyze the meanings of the search queries. The content synonymically linked to the search query, are also returned. Google mainly carried out this to fight spammers trying to cheat the search engine by using keyword stuffing.
What are LSI keywords?
LSI keywords are the keywords that are semantically related to the keyword you are searching for. This is done possible by Google making use of Latent Semantic Indexing.
Have you ever wondered how Google and other search engines differentiate between homonyms (same word with different meanings)?
You may have noticed this. Google highlights the synonyms in SERPs.
Let me take an example. Search for “ps picture size”. Google highlights the words "picture, photo, image". Sometimes the synonyms highlighted in SERPs are also called Synonymous LSI Keywords (sLSI). Now you got a clear idea what are LSI keywords.
The term “PS” has different meanings in Google like PlayStation, Photoshop, and so on. Google gets results based on the keywords you have used after the term “PS”. Consider an example, where I search for “ps image size”. Google comes to know that the term “ps” in this context refers to Photoshop. Because I have used “image size” in front of it. Now Google searches all the content containing the terms Photoshop and image size, both.
By doing this, it can get what the user is trying to search. This is a part of Google's LSI algorithm.
Here's the kicker.
Here once I append the word "fps" to the "ps" keyword. Google easily made out that I'm talking about either first-person-shooter games or the frames-per-second in PlayStation devices.
So, you now know that the FPS is a semantically related keyword to PlayStation.
Google determines the meaning of the content by looking at the other semantic keywords that appear along with the base keyword and establish a relation among them.
What is Latent Semantic Indexing?
LSI algorithm helps Google to understand the content on the web page like we humans do, and match it with the search queries of the user. This results in more accurate search results for the queries of the users.
LSI also helps search engines to determine the semantic relationships between two or more documents.
Semantically close: When two documents have large number of common synonyms, similar phrases, phrase correlations, common entities with close proximity, then they are determined to be semantically close.
Semantically distant: When two documents have few common words, synonyms, correlations, they are determined to be semantically distant.
Based on these factors, a similarity value is calculated that helps Google to return the most relevant result to the query of the user.
Likewise, search engines can determine whether the user’s search query and the documents are semantically close by looking at the common synonyms, context and related entities to that of query and the documents.
The bottomline is that, if your webpage has synonyms, related words, entities related to the main keywords users are searching for, chances of your webpage ranking high for the keyword is higher. These keywords that you need to include in your webpage to get a LSI boost are called LSI keywords.
These are some of the ways by which you can get LSI boost.
How to Find LSI keywords?
You now know about LSI keywords. They are the related keywords for a particular keyword. There are several LSI keyword tools. Among them, the Google search is an excellent tool. All that matters is, you have to learn how to use it. Here's the simplest way. You can find LSI keywords from search suggestions. And also from related searches at the bottom of the SERPs.
Some of the overlooked LSI keyword tools
These Related keywords are also called Related LSI Keywords (rLSI). Apart from Google search, you can also use the Google Keyword Planner. The goal is to note all the synonyms to your main focus keyword.
Another tool that you may consider using to do the LSI keyword research is the best visual keyword research tool called - AnswerThePublic.
It helps you easily make out relations with various entities and come up with less competitive LSI keywords you can target.
You can also make use of a tool called LSIGraph. This is a tool that's specially meant for LSI keywords research. Although there are many such tools, the fact that this tool is named after the buzz term "LSI keywords" contributes largely to it's popularity in the SEO world.
The results returned by the LSIGraph are LSI or related keywords. You can upload these keywords in the free bulk keyword volume checkers like KeywordsEverwhere to get an idea.
Getting LSI Keywords That Move The Needle For Sure
While researching LSI keywords, knowing what of these keywords will make a difference to your rankings is a challenging task.
That's where you need to do the is use the TF-IDF concept.
What the heck is TF-IDF?
TF-IDF stands for term frequency-inverse document frequency. It is used by search engine algorithms to determine the relevancy of the web content for a specific query. The search engines treat those keywords that are present alongside the main keyword in many documents (that target the same main keyword) as related/LSI keywords.
Along with this, search engines also learn how often a LSI keyword will ideally be present in document on a specific topic.
There are free tools like Website Auditor from SEO Powersuite, that can help you do the TF-IDF analysis of top ranking websites. By doing this, you can know what LSI keywords your SERP competitors are using and how often.
If it works for them, why not you?
As you can see in the above screenshot, the tool has scraped the content from top 10 search results for the keyword - "LSI keywords". It has made a huge list of multi word keywords and also single word keywords that appear prominently in the results.
The frequency at which these terms appeared in the documents is referred to as TF-IDF. It also displays the TF-IDF of your page you want to semantically optimize for. You should always aim for reaching the average search results' TF-IDF in the webpage you need to optimize.
If you are witnessing a new keyword your competitors are using, but you have a TF-IDF score of 0 for that, then it means you need to consider updating your webpage and include the new concepts/keywords accordingly.
Many newbies ask me, what keyword density is ideal for an article. But the answer is that the density of the keyword varies greatly depending upon the topic. So, the best bet for you is to take a look at what other blogs that are ranking high for the keyword is doing.
Rather than stuffing the same keywords repeatedly in your articles, follow what's working for others already.
Rather than doing the LSI keyword addition after publishing your article, which is a real pain, do research beforehand.
Before even starting to write an article, prepare a solid outline of the article. Along with the outline, research the main keyword and also the LSI keywords you need to include in the article using competitor keyword analysis.
Adding the LSI keywords after publishing the article is quite irritating and as well inefficient job.
Tapping Into The Search Console For LSI Benefit
Google search console displays you the keywords that are driving traffic to your site. Earlier, Google Analytics was displaying it. But, now search console is the only place where you can get keywords that are driving you traffic.
To get the keywords that are driving you traffic, go to Search Console, and there head over to Search Traffic and Search Analytics.
1. Take a popular blog post, now dig in the search analytics to come across what are all the related keywords that are driving traffic to the post alongside the main keyword.
2. List the keywords that are driving traffic to your page inspite of you not targeting it. Now, go to your page, copy the main keyword you need to rank for. Now search for that keyword on Google.
3. Browsing the top 5 SERP results for the main keyword, notice how often they've used the related/LSI keywords you found in the first step above. Tweak your post and add the necessary information along with the LSI keywords to increase your chances of outranking your competitors.
Dig in competitor research tools
Here's another strategy I'm using right now. Once I come with a target keyword to create a blog post on, I'll plug the keyword in Google search and take a note of the top 3 ranking pages for that keyword.
Now I make use of tools like Ahrefs and SEMrush to know what are all the other related keywords that are driving traffic to these top 3 results.
For example, if a site is ranking for the popular keyword - best turntables under $100, it will also be ranking well for best record player under 100, best budget turntable, best cheap turntable, top turntables under $100, best turntables below $100, and some other synonymous variations. The above mentioned tools like Ahrefs, will return all these keywords that are most of the times LSI keywords.
LSI keywords to avoid
Don't use LSI keywords in your article just for the sake you need to use it. Make sure that the LSI keywords that you are using are within the context of your content.
If you include an LSI keyword just because of the fact that it's related to the main keyword, and do not include relevant information related to it, it will do more harm than good.
It's easy to lose the track when you go bonkers about making your webpage semantically and entity rich.
Before including any LSI keyword make sure that your article is useful for also those who search for that LSI keyword alone in Google.
In short, by including LSI keywords and not including relevant content accordingly, you still hurt topical relevancy of your webpage.
How to use LSI keywords for SEO?
Let me give an example. You are writing an article on Complete Guide to Recover From Google Panda Update. Your prime keyword would be "Recover from Google Panda Update". LSI keywords are "Rescue your site from Google Panda Update, Recover from Google Panda effect, etc."
After you write a post on this, make sure you include related terms like effect, SEO, Penguin Update. If you have a title - “Recover from Panda”, and LSI keywords like SEO, update, etc. in the post, it acts as a content relevancy signal. Once you put terms like "animal, attack, white bear, etc." Google thinks that your post is about recovering from Panda animal attack. Avoid repetition of the same keyword over and over.
As you can see in the above screenshot, Ankit in his post - Blogging Tips, has made clever use of LSI keywords for targeting the main keyword - "Blogging tips for beginners".
If you are interested in knowing more about LSI or semantic keyword research, do refer the ultimate guide highlighted below.
Must read: The Best Guide to Semantic Keyword Research
I've made a separate post on some of the greatest tools for LSI keywords for the research.
Where to include LSI keywords?
LSI and other keywords follow some basic rules. You can include LSI keywords anywhere in the article. Below is the list in the order of priority,
- title tag
- meta tags
- header tags
- anchor texts
- beginning paragraph
- body of content
- ending paragraph
Even, stuffing the LSI keywords all through the blog post smells spammy.
Gone are the days when Google was looking for the number of times a keyword is used in an article. Now Google needs to find your content contextually relevant and it expects some keywords to be included in your content to measure the relevancy.
"Using LSI keywords in SEO-friendly and natural manner is better than using the same keyword often in your blog post".
As I said before, use related keywords only to signal content relevancy.
Rather than using the same keywords repeatedly on your site, once you hit that limit, start including LSI keywords. With this, your content appears perfectly natural and as well optimized.
If you are outsourcing your blog posts to a writer, make sure that you give him the LSI keywords along with the main keyword to use in the blog post.
Set the clear expectations, by presenting him a complete outline of the blog post, keyword usage, and questions he need to answer in the blog post.
Teaching your freelance writer about how to use these keywords in the article makes a great difference in the SEO traffic you get.
Hummingbird and LSI Keywords
In 2013, Google rolled out Hummingbird update. Some also call it as “Semantic Update”. This is an algorithmic update aimed at understanding the intent behind an users’ search query in a better way using semantic intelligence.
Instead of making out the meaning of individual words in a query, it makes out the context and meaning of the whole query. This proved to be very helpful for Google to improve its voice search engine, results returned for new search queries, and much more.
Once Google starts to know the particular intent behind a query tapping into its big database of entities, it started to expect some keywords in the webpages to make sure that they are relevant to the search query.
In other words, prior to Hummingbird update, Google most of the times was matching only keywords in the webpages and displaying them (that’s when keyword stuffing was abused a lot). After the update, Google started showing the relevant web-pages to the specific intent of the keyword.
The update impacted 90% of the search queries.
After this update, some authority sites that did not have keywords present in the article but indeed highly relevant to the search query started appearing. It resulted in top 5-10 results getting reduced clicks.
After the HummingBird update, the term topical optimization came into picture over keyword optimization. Topical optimization refers to optimizing your article for a narrow topic than a particular keyword. For this you need to make use of LSI keywords and answer every possible question regarding the article you are writing about.
LSI keywords for reviving dead content
Every blog has content that old and stale. Those blog posts that are getting little or no Google love.
It’s time for you to show some life for those blog posts with some LSI magic.
For this, you need to go to Google Search Console.
Here you need to sort by descending order of impressions.
As you can see in the above image, this very article is also appearing frequently for “LSI graph”. It is a tool for researching LSI keywords. If I include this tool frequently on the blog post, then I will rank higher.
Another trick is to make use of SEMrush.
Above are some of the LSI keywords for my main - “LSI keywords”, that I am ranking for in top 50. But, frequently using them in my post makes me gain some additional traffic.
Like this, by making use of LSI keywords you can drive some additional traffic to your old blog posts that would otherwise drive you very low traffic.
LSI keywords for PPC effectiveness
If you are doing PPC ads, don’t just target a single keyword. You can research good related keywords to your main keywords and increase the chances of lowering the costs and maximizing the visibility of your ads.
LSI keywords to the main keyword typically has low competition. So, the CPC will be low and you get good ROI for your advertising efforts.
You need to think outside of the Google Keyword Planner, to truly find out the hidden gem keywords you can target. This strategy not only applies for Google Adwords. You can use the same in Bing Ads, Amazon ads, and all PPC networks.
Sometimes, the combined traffic from these LSI keywords will be higher than that of main keyword with less CPC rate.
If you stumble upon a keyword with high competition, research for LSI keywords and target them instead of the high competition keyword. In doing so, you need to ensure that the LSI keywords reflect the same intent of your ad. Or else, it’ll backfire.
LSI keywords for image optimization
It’s been advised to use the keywords in your image name and alt tags. But do you know that by using the same keyword in all the images on my web pages I hurt my organic traffic?
People often think that, if they stuff keywords in image alt tags and names, it’s ok. They also think that using the blog post title as alt text for all the images on a page is fine.
But my experience tells me that alt tag keyword stuffing pushes your webpage to 8th or even 50th position that would otherwise deserve 2nd or 3rd position.
Instead of using the main keyword, try to make use of LSI keywords in your image alt tags. Don’t use them if they are irrelevant to the image. Try to supplement them with image description. ALt text should be highly relevant to the image.
LSI keywords in your backlink campaigns
Getting the exact-keyword anchor text backlinks helps you to rank.
But how much is too much?
Google is frowning upon keyword-rich anchor texts lately. So, there are several fixes for this.
Brand name backlinks
Naked URL backlinks
LSI keyword backlinks
If your backlinks’ anchor text are not too optimized, you significantly reduce the risk for penalty even if the backlinks are from low quality sites or a content farms.
As an alternative to using main keyword as an anchor for all the backlinks you build, try to make use of LSI keyword anchor texts.
Suppose you want to rank for “furnished apartments in seattle”, you can build backlinks with the following LSI-rich anchor texts.
furnished housing in seattle
serviced apartments seattle
furnished flats in seattle
But still, keep the keyword anchor backlinks percentage below 30-40% in your backlink profile depending upon the quality of the sites you’re getting links from and also the niche.
Want to use this infographic freely on your site? Just copy and paste the embed code below:
I think you have liked this post. If that is true, do share this post and speak up in the comments section.