Are you paying enough attention to your on-page SEO? I bet you’ve done your homework with keyword research and keyword density but how about LSI keywords?
If you haven’t heard of the latter or don’t know much about it now is the right time to improve your knowledge and start using LSI keywords as an integral part of your SEO strategy.
A short note before we dive in.
This article won’t discuss topics like off-page SEO, white hat link building or even keyword research.
Instead, it will be dedicated entirely to one of the less known aspects of SEO – LSI keywords.
Today you will learn how to improve your Google rankings by using LSI the right way.
For your convenience I created quick links, so if you are interested in a particular chapter of this article just use them.
Table of contents:
- Keyword Density – the predecessor of LSI keywords
- Keyword stuffing – the dark ages before Google Panda existed
- Does keyword density still matter?
- LSI keywords – the modern way of doing on-page SEO
- How to find LSI to rank high in Google?
- How to use LSI keywords in practice?
So without further ado, let’s jump right in.
Keyword Density – the predecessor of LSI keywords
If you’ve been in the SEO industry for a few years now you have definitely heard the term keyword density.
Simply put, keyword density shows the number of times a word appears in your text. It is measured in percentage. The bigger the percentage, the higher your chances of ranking better in Google for your preselected keywords.
At least this is how it used to be before Panda and Hummingbird. Those were simple times.
You see, back in the day keyword density was THE only way in which Google search engine could define if a particular page is relevant to the query the visitor is searching for.
Needless to say people started taking advantage of this metric pretty quickly.
That’s how keyword stuffing was invented.
Keyword stuffing – the dark ages before Google Panda existed
I bet that almost everyone who’s been trying to optimize their web page for a particular keyword has tried keyword stuffing at least once in their lives.
In a nutshell, this is the process of adding your main terms more times on your web page so that you start ranking higher in Google for them.
Makes perfect sense, right?
And it worked… for years and years.
After all, besides the links anchor texts (where the concept was pretty much the same by the way) keyword density was THE most important on-page metric out there.
You just had to include your main keywords in your page’s title, META description, META keywords and to spam the hell out of your body of content and BOOM! You just boosted your site’s rankings big time.
And I’m not talking about small jumps. You could literally enter top 3 in Google just by increasing your keyword density by 10-15%.
Doesn’t sound too good anymore, eh? No.
In order to start cleaning their results, Google had to find a better way to evaluate the relevancy of each webpage on the Internet.
This long and tedious process began with the launch of Google Panda algorithm in 2011 which was additionally improved in 2013 when Google Hummingbird was initially launched.
In a nutshell, Google Panda tackles sites with low-quality, duplicate or thin content, while Hummingbird is designed with the main purpose to better focus on the meaning behind the words (semantic search).
With these two important SEO updates, Google pulled the switch on a huge chunk of useless, low-quality and spammy content which took the game to a brand new level.
In 2016 Google launched their newest algorithm which is based on machine learning. It was called Google Rankbrain and was created in order to help Hummingbird with understanding the ambiguous search queries.
Today, I would recommend that you never exceed 2% of keyword density for your keywords. This way you will stay away from Panda and keyword stuffing problems.
You’re probably wondering if Google used LSI keywords before Panda.
My assumption is that Google has probably considered LSI for over 10 years now. Probably even from the very beginning.
But we probably won’t be mistaken if we say that it started to pay more attention to these specific keywords in the past 5 years or so.
Nowadays, LSI keywords are considered one of over 200 Google ranking factors that every SEO expert and webmaster should have in mind in order to improve their SERP rankings.
Does keyword density still matter?
Five years after Google Panda was initially launched, we can safely say that keyword stuffing has become an obsolete technique which not only doesn’t work anymore but can even worsen your SEO.
OK, but how does this apply to keyword density? Should we still pay attention to it?
Yes, you should. I will explain why.
Let me give you an example with the term that I’m targeting with this article – LSI keywords.
I analysed the top 10 ranking websites for this search query and checked the keyword density of each of them. You can use Free Keyword Density Analyzer Tool from SeoBook.com for this purpose.
Here’s a table with the results in terms of keyword density percentage (click on the image to enlarge):
After my keyword density analysis was completed I already knew the following four very important things:
- The most often used keywords for each of the top 10 players for the term (LSI, keyword, keywords, search, SEO, Google and 10 other keywords that are shown in the table above)
- The average percentage for each of them on each of the top 10 sites
- How many times each of the main terms are being mentioned. Naturally, terms like “LSI” and “keyword”, which are present in 10/10 of the cases, are the words with the highest correlation level. This means you MUST use them too.
- The average content length for each of the pages. The longer the average page content is, the longer content you should write yourself. Content length does matter big time.
This is important!
If you want to stand a better chance of ranking among the top 10 competitors in your niche you should always perform such LSI and keyword density analysis upfront. This way you will know what you’re up against and how to outrank your main competitors in Google.
OK, back to my LSI example.
After my table was completed, I knew the average percentage of each of the top ranking pages in Google. In other words, I already had the keyword density range I needed to follow.
Now, from time to time you will come across sites that are showing much different keyword density levels.
Look at LSIGraph.com, for example.
Obviously its keyword density percentage is quite high, so we shouldn’t take it into consideration and stick to the average percentage of the other 9 websites, that is 0.7-2.8% for “LSI”.
Repeat the same procedure, excluding the most irrelevant results from your final list.
Remember, SEO is all about finding patterns, analyzing them thoroughly and getting the best out of them. It’s a game in which you must always keep pace with the best players in order not to sink into oblivion.
With that being said, keyword density analysis is an inseparable part of the whole process. It gives you invaluable insights over the best ranking sites in your industry and how they do their on-page SEO.
Simply put, you have to know which keywords your main competitors are using and what their percentages are so that you can replicate these stats and fit in perfectly.
NOTE: These techniques should always go hand in hand with your link building efforts, but as I mentioned in the beginning of this article, we’re only interested in the on-page stuff right now. Also, you should always try to write high quality and useful content, otherwise you cannot expect to rank among the best sites in your niche.
LSI keywords – the modern way of doing on-page SEO
So far we covered all the main aspects of search engine optimization from on-site perspective.
You already know that keyword density still matters, while keyword stuffing doesn’t.
Now it’s time to take your SEO to another level by diving into the world of LSI keywords.
What does LSI keywords mean?
LSI stands for Latent Semantic Indexing. It might sound too complicated at first reading, but it’s not really.
LSI keywords are simply closely related words or phrases that are semantically related to each other.
Does this make them synonyms? Not exactly!
Latent Semantic Indexing keywords are not just synonym words. They are words that often go hand in hand without being direct synonyms. I will give an example about this in a bit.
Let me explain this with a real LSI keywords example.
Have you ever wondered how Google understands what a particular web page is all about?
Let’s say that you are writing an article about cars.
You may well be writing about one of these 5 things:
- Cars – the vehicle
- “Cars” – the animation movie
- CARS – the Canadian Association for Rally Sports
- CARs – the Canadian Aviation Regulations
- The Cars – the American 1970s music band
So how does the search engine algorithm know what topic you are writing about? (have in mind that Google does not recognize capital letters!)
Exactly! By using LSI keywords.
Basically, the Googlebot scans your page searching for closely related terms which can help it understand your page content better.
Back to our LSI keywords example about cars.
Here are some Latent Semantic Indexing keywords that Google would have found on your pages if you were writing about one of the topics given above.
- Cars – the vehicle (vehicle, used, new, buy, sell, dealers, repair, etc.)
- “Cars” – the animation movie (film, movie, produced by, directed and co-written by, motion picture, Walt Disney, etc.)
- CARS – the Canadian Association for Rally Sports (association, rally, sport, championship, Canadian, etc.)
- CARs – the Canadian Aviation Regulations (aviation, regulations, administration, aerial, aerodromes, airports, etc.)
- The Cars – the American 1970s music band (rock, band, music, guitar, drums, song, etc.)
See, that’s how Google knows!
And to dig into this example even deeper (in case you feel nerdy today 🙂 )…
There are actually lots of other fancy sounding terms around Latent Semantic Indexing. Here’s a list with some of them:
- Latent Semantic Analysis
- Natural Language Processing
- Phrase-Based Indexing and Retrieval
- Term Frequency-Inverse Document Frequency
- Latent Dirichlet Allocation
- Probabilistic Latent Semantic Analysis
- Hidden Topic Markov Model
- Latent Dirichlet Allocation
Sounds difficult? Well, it kinda is. But you don’t have to browse all of them in Google or Wikipedia and try to become semantics experts.
As Cyrus Shepard, ex-Mozzer, puts it:
Keep in mind these are only some of the ways in which Google could determine on-page relevancy, and they aren’t absolute law! Experimenting on your own is always the best policy.
1. Keyword Usage
3. Synonyms and variants
4. Page Segmentation
5. Semantic Distance and Term Relationships
6. Co-occurrence and Phrase Based Indexing
7. Entity Salience
If you want to get deeper insights on these 7 points, I suggest that you read this great article of Moz which explains this difficult terminology.
OK, back to practice…
How to find LSI to rank high in Google?
No more theory, let’s dive into the practical stuff.
There are 4 easy ways to find Latent Semantic Indexing keywords:
1. Google’s instant search
2. “Searches related to” results
3. Bold words in SERP results
4. Keyword research tools
Let’s break them down step-by-step:
Find LSI keywords via Google Instant Search
The first way you can identify LSI keywords for your SEO campaign is by simply writing your keyword into Google’s search box and see the 10 suggestions it returns for you. Remember to turn on your instant search feature in your browser in case you have previously turned it off, otherwise this method won’t work for you.
Find LSI keywords via the “Searches related to” feature
Just type your keyword in the search box and scroll down to the bottom of the results page. There you will find 8 additional suggestions you can add to your LSI arsenal. Note that some of them might not be a nice fit so make sure you analyze them from human perspective for best results.
After all, you don’t want to include LSI words like “fun easy learn English” in your article despite the fact that Google tells you to. It’s simply not natural. Instead, you may consider changing your LSI keywords into “fun and easy way to learn English”. Much better!
Find LSI keywords via bolded words
When you perform a search you’ll notice that Google shows certain words in bold text.
Do you know why?
Because it considers these words closely related to your search string. In other words, these are considered LSI keywords. Whenever you notice some keywords in bold make sure you implement them in your own copy. They will help you make your page more relevant to your subject.
Find LSI keywords with the help of some tools
If you have already exhausted the first 3 methods for finding Latent Semantic Indexing keywords it’s a good idea to enrich your LSI arsenal by using some LSI tools.
I suggest that you begin with the most popular tool namely, Google keyword planner.
Here’s how to do this.
Firstly, put your keyword through the tool and click “get ideas”.
You will see two different tabs in this tool which are great sources for LSI keywords:
Ad group ideas – this is a much underestimated option, but is one of the best sources for LSI keywords ever. Just go through the list with suggestions and make sure you include as many as possible in your final copy.
Keyword ideas – everyone knows this one but very few of you have actually used it as a way to find LSI keywords. You can start using it today.
Other great ways to scrape for LSI keywords is by using some small SEO tools and plugins specifically designed for this purpose.
LSIGraph tool – LSI keyword generator
If you run a search for LSI keywords in Google you won’t miss this one. It’s ranking in top 3.
LSIGraph is a very easy to use, but yet very powerful tool. Just put your desired keyword in the search box and hit “Generate”.
The LSI keyword generator will come up with a long list of potential LSI keywords that are a perfect fit for your string. You will be pleasantly surprised how many more keyword suggestions you will get from this nifty free tool.
As its creators said it, SEOPressor is an on-page SEO plugin for WordPress that helps you do on-page search engine optimization.
It’s pretty simple to use too.
You can learn all about how to get 100% from the video below:
SEOPressor also does multiple keyword analysis, over optimization check, social SEO integration and structured data support. Finally, you can even do an automatic smart linking to reduce your bounce rate.
This is a paid LSI tool, but its cheapest plan starts from as low as $5 per month so it’s quite affordable.
Another good WordPress plugin for LSI is Twinword SEO Plugin. This is LSI, medium tail and long tail keywords research tool, that automatically suggests keywords related to your blog post while you write with each word’s monthly search volume, number of results, and KEI.
Its features include: auto keyword suggestions, monthly search volume, number of results and KEI (keyword efficiency index) and keyword usage tracking. You can also sort the information via filtering.
There are multiple other keyword generators which can be used as LSI keyword generators, but here’s a list with my 4 favorite ones:
- SEMRush related keywords
- Übersuggest (show only versions around your root keyword)
- KeywordTool.io (show only versions around your root keyword)
I’d like to say a few words about the Keys4Up tool and about the SEMRush related keywords feature. This will throw some additional light over the whole LSI concept.
As I mentioned earlier in this article, LSI keywords are not exactly synonyms. They actually are closely related and relevant terms!
Let me describe this better with yet another example.
LSI keywords example or why they are not just synonyms:
Imagine that you want to write articles about pizza and you want to know which terms you have to include in order to stand a better chance to rank high with LSI from on-page point of view.
Run the word “pizza” through Keys4Up first.
Now let’s see what SEMRush related keywords tool holds for us:
See the HUGE difference in the results?
We’re no longer seeing a list with keyword ideas based solely on the word we enter in the tool (like with Übersuggest and KeywordTool). Instead these results give us a completely new understanding on Latent Semantic Indexing keywords.
If you’re writing about pizza and you want to make sure you are using LSI to rank high the right way, make sure to also include terms like: cheese, pepperoni, delivery, crust, food, ingredients and restaurant for better results.
LSI keywords are not just variations of your main keyword. They are closely related terms which Google likes to see when it crawls pages about a particular keyword.
It’s all about patterns.
How to use LSI keywords in practice?
Now that we already know what LSI keywords are and how we can find them, it’s time to start using them to improve our on-page SEO and organic rankings.
OK, so how do we use LSI to rank high in Google?
You have already come up with your list of Latent Semantic Indexing keywords and now is the time to start integrating them in your web page.
Here are the main areas where you can implement LSI keywords to get the most out of them:
- Page title
- H1 and H2 tags
- URL address
- META tags
- Images alt text
- First paragraph of text
- Body of content
- Links anchor texts
- Last paragraph of text
By sprinkling LSI keywords through each and every one of these elements you will surely get the best from your organic SEO.
If you prefer to ask a professional copywriting expert to help you with your content writing, I recommend the services of my friends at ProfCopy.
Before you go you might want to check this ultimate guide on LSI from by friend and colleague Hussain Omar
Now you know pretty much everything about the Latent Semantic Indexing keywords concept and how to use it to skyrocket your website without a single link!
I hope you find this article useful and it becomes an integral part of your on-page, marketing and keyword research strategy.
If you have any questions, don’t hesitate to contact me via Facebook or to send me an email 😉