If you ever get a chance to read about Latent Semantic Indexing (or LSI), some might think you would need to be at MIT to understand what is it about.
But most would probably think, “eh?”
Don’t worry if you are on the latter. It is actually a fairly simple concept but made complicated because of various names (like Latent Semantic Analysis) and the different definitions that you can find online.
Even reading a chapter about it from a university-published book can be quite intimidating.
You can get very deep and analytical on this topic but let’s not, because most of it is not relevant to us.
And ironically, the idea of staying relevant to the topic is very similar to the topic of Latent Semantic Indexing.
What is Latent Semantic Indexing?
To put it simply, LSI is a “mathematical method used to determine the relationship between terms and concepts in content”.
The easiest way to define it is to think about how some keywords are related to your main chosen keywords.
So what does it have to do with your website?
Well, a heck of a lot.
When Google crawls a specific web page, the most common words and phrases are collated together and end up having related meanings.
The purpose for LSI is to allow Google’s crawlers to to check out your content, learn about the possible synonyms related to your content and what “categories” to place it in.
Let’s look at a group of Latent Semantic Indexing example to give you a better understanding.
Example 1 – Pizza
Imagine a group of people who are discussing where to eat and they want to go out for a pizza.
You can appreciate that there are different ways of describing a place where you can eat pizza based on where you’re from, the local slang, colloquial vocabularies, etc:
- “Pizza restaurant”
- “Pizza place”
- “Pizza joint”
- “Pizza takeaway”
- “Pizza delivery”
LSI essentially helps to understand that each of the keywords above are related to the same thing.
In this case, it is a place where you eat pizza as opposed to e.g. pizza recipes.
And you only have to search for “pizza place” to see how Google is picking up the related words:
Notice how Google has made certain words in bold, like the word ‘restaurant’?
They know it’s a related keyword to “place”.
Example 2 – Windows
If you are searching for “windows”, there are hundreds of related keywords that you can think of:
- “Bill Gates”
- “Windows 10”
- “Surface tablet”
These keywords are naturally grouped together and rightly so as these are the potential LSI keywords when writing a post about “windows”.
And LSI helps to differentiate from the “other” windows:
- “Window cleaning”
- “Double glazed windows”
- “Wooden windows”
- “Window locks”
Example 3 – Cars
Imagine if you are searching for information about cars.
The way you search for cars could tell Google a lot about you and what you are looking.
But the topic of “cars” could be about anything:
- Cars – the actual vehicle (the type of cars, the brands available, new or second hand, the mechanics, etc)
- Cars – the Disney movie (all 3 of them)
But if you search for “self driving cars”, it will unlikely bring up the typical results that you tend to see when searching for “used cars”, “BMW saloon” or “electric cars”.
Those keywords tend to bring up official dealers, local advertisements or companies which sell these cars.
But at the time of writing, “self driving cars” are bringing up magazines and news articles as it is a very niche topic to discuss.
LSI has helped to do that.
And notice how keywords such as “Toyota Prius”, “Audi TT” and “Lexus RX450h” are in bold?
Evidently, they are being used as self-driving cars as you can see in the meta description.
So yes, LSI has helped to do that.
The Benefits of Latent Semantic Indexing
One of the key areas that Google has struggled with is understanding the fact that multiple words can have similar meanings (or synonyms) like the pizza or car examples above.
The problem is that search engines are not human, so it struggles to understand and naturally process the words and languages used like us humans would.
But that’s where LSI comes in.
This is the very key which helps Google to understand the relationship between certain words, as close to human beings would do as possible.
Scary stuff, right?
But LSI is a good thing because it helps to keep content relevant, high quality and provide benefits to your visitors.
And it will also provide benefits to you, the creator of the content who is, in some way or form, forced to create quality material.
In the old days, marketers could get away with abusing Google’s result page by keyword stuffing and other black hat SEO methods.
But you can put to bed your plan of keyword stuffing your content. Or writing just a few hundred words with a hope of attracting organic traffic.
It just won’t work anymore.
Now, you have to be creative with your keywords.
It’s good for you, good for Google but more importantly, it’s really good for your readers.
How to Find LSI Keywords Using Google?
Google is becoming smarter and smarter at understanding your content, so it doesn’t matter how many times you cram a specific keyword in the article.
But LSI keywords (or Latent Semantic Index Based Keywords) can better help Google to understand what your page is about.
We’ve mentioned earlier about how Google bolds related words that are very similar to your chosen keyword.
Those are your LSI keywords.
The idea is to find the relevant LSI keywords and add them to your content.
But finding them is easier than you think and I will show you 2 very easy ways to find them.
One place is at the bottom of Google, named “Searches related to…”
After you search for your primary keyword, scroll to the bottom.
This section is a goldmine to find related keywords, so find any keywords that make sense and slot them in your content.
This is a great place to find long-tail keywords that you would not have been able to find on Google’s Keyword Planner.
In fact, I have used the very same technique for this blog post that you are reading.
It’s not a surprise that Google’s search result page is a fantastic place to find content and keywords ideas.
How to Find Related Keywords Using LSI Keyword Generator?
Another easy way of finding related keywords is to use an LSI keyword generator, and there are 2 of them which stands out.
One of them is called LSI Graph; a powerful and free tool which generates a list of keywords relating to your primary keyword.
It couldn’t be simpler; type in your chosen keyword, do the annoying captcha, click on ‘Generate’ and there you have a long list of options to choose from.
The added benefits of using UberSuggest is the useful feature of filtering whether you want a list of keywords found on the web search result, images, shopping, news or even YouTube.
You can also filter the result further by choosing your preferred geographic target as well as download the results to your computer.
Using Latent Semantic Indexing for Content Ideas
It might have become obvious for you but I’ll say it anyway.
When researching for LSI keywords, they can also help you find relevant content ideas which are more likely to satisfy what your audience are searching for.
You can even use it to add to your existing content too.
For example, if you specialise in writing about office furniture, you might have separate sections in your existing blog post that covers different types of furnitures.
Or you can go deep, in depth and focus on particular type of furniture instead of looking at it all in one place.
How to Use Latent Semantic Indexing?
LSI is not something to be worried or paranoid about.
And there are no tricks or clever schemes you need to think of to get the full benefits of it.
It’s actually very common sense really:
- Make sure your content is relevant to the topic. If the title is about cake recipes and you write about cookies, that’s not relevant.
- Be wary of overusing keywords or writing with poor grammar. If you read your content out aloud and it doesn’t make sense, then it’s back to the drawing board.
- You cannot automate a blog post. The best content out there are created by hands with blood, sweat and tears.
- If you prefer to outsource your content, in most cases, you get what you pay for. So if you think you can acquire high quality via the likes of Fiverr, you need to wake up.
- Check your Google Analytics and look at which of your blog post are popular. Could you improve it by expanding on it, adding more images, thinking of more LSI keywords. Could you even
- write another blog post in similar style to your most successful posts?
- Use Google Search Console or tools such as SEMrush to understand what keywords your pages are ranking for. Is it relevant to your post? Are you missing any potential LSI keywords?
If you follow the natural process of creating a quality, relevant and rich content, more or often than not, Latent Semantic Indexing has already being applied.
Even though it’s worth being aware of LSI, it’s not something that you should force.
Allow it to placed naturally in your content.
But it’s important to be aware of the searcher’s intent behind your primary keywords.
As much as it’s tempting to add a keyword with high search volume, if it doesn’t match the searcher’s’ intent, you are wasting your time.
For example, don’t write a blog post which covers both “chocolate cake recipe” and “chocolate cupcake recipe”.
Those who are searching are more likely to be interested to learn more about one of the two topics rather than both of them in the same place.
So have a go at finding synonyms for existing or new content.
Not only will you improve your keyword researching skills but it can widen your visibility on search results and potentially attract more traffic to your site.
And who doesn’t want that to happen on your site?
So how do you feel about Latent Semantic Indexing? Do you understand it better? Did you know that such thing even existed?