Lately we have been really upset that search engines do not quite get what we want from them. The main factor here is the search for semantics and its infinite possibilities.
Now it can give an answer right in the search page:
What do you think about this? The excerpt shortly thereafter became prominent as below:
Outstanding result excerpts of Bizzvn
In the context of Content growing exponentially, I decided to explore the problem and find strategies to optimize semantic search. And Google is focused on semantic search because I am sure this search engine is really the best among search engines right now.
How does semantic search work?
Machine learning versus Artificial Intelligence
How to optimize semantic search?
- Create entities
1) Apply structured data to the website
2) Become a Knowledge Graph entity
- Pursuing search purpose
1) Create semantic keyword groups
2) Check out the keywords to be relevant
3) Develop content according to user intent
4) Optimized for voice search
- Create topic groups
1) Identify Pillar and Cluster content pages
2) Check your site structure
3) Manage your internal links
- Amplify your content
1) Do case studies and surveys
2) Add useful visual information
3) Available on Social media
4) Get customer reviews
The danger of AI
What is semantic search?
Semantic search or semantic search is how search engines handle natural language and understand the search intent of a query through contextual meaning to provide relevant, personalized results.
Usually, when people think of semantics, they tend to think about the meaningful area. However, there are several types of semantics. Below is an outline of what is important to our investigation:
- Logical Semantic refers to the relationships between linguistic concepts / elements (For example: references, assumptions, implications of words). In SEO it can (and should be) used to structure content. For example, structured data plays a big role in logical semantics. But not only, logical semantics are also used to build websites, meaning it is found in the architecture of HTML and web pages.
- Semantic Lexicals communicate with the meaning of words and their relationships. In SEO it is used to do keyword research.
How does semantic search work?
The pursuit of semantic search of search engines is understandable. First of all, that means less spam, better natural language processing and better understanding of search intent, leading to the best user experience. Secondly, it is a known fact that data is doubling every two years, so the demand from search engines needs to better organize this growing structure and data connection.
The entire content with semantic search begins with a Hummingbird update which was released by Google in 2013. This algorithm uses context and search purpose (not individual keywords in queries as before) to make sure the pages that match the meaning work better than the pages that exactly match the keywords. It is a fundamental change in the way Google distributes its results and ensures a satisfactory search experience.
Then, in October 2015, Google launched RankBrain as part of Hummingbird. Although its purpose is similar to Hummingbird, it works differently. RankBrain is a machine learning system that consists of two components:
- Query analysis – it tries to interpret queries by combining them with more common queries. The learning process is triggered when the algorithm encounters long-tail queries, which are unclear or unfamiliar.
- Ranking – to find a good match for a query, this component analyzes pages that have been indexed for specific features, for example, usage patterns of certain related terms. These specific features are determined by analyzing the best performing search results (by CTR, bounce rate, time on page, etc.) and looking for similarities between these pages. Therefore, pages that are identified as responding well even without an exact word from the query, are considered relevant.
Machine learning versus Artificial Intelligence
Machine learning is the science that makes computers come to conclusions based on the data they have but are not specifically programmed to perform those tasks. In other words, it is a mathematical operation system to solve problems. Machine learning is a subfield of AI. People often vaguely use terms like “machine learning” and “artificial intelligence”. While combining these two concepts is normal when it comes to RankBrain, we should understand that machine learning is not AI. To avoid any misconceptions, find out what the two terms mean in nature.
Artificial intelligence is aided by machine learning. It is the science of creating systems that have process and human-like information in the corresponding way. It means that systems operate in a creative and less predictable way, just as humans do when interacting with anything in their lives (they can either correctly guess or be completely non-existent). physical).
There are three AI classifications:
- Artificial Narrow Intelligence (ANI) – it is the AI that solves a specific task in a way that compares to a person’s capacity or exceeds them (for example, eliminating spam);
- Artificial General Intelligence (AGI) – that is AI that can perform any task. When AI can perform like a human, it is considered an AGI;
- Artificial Superintelligence (ASI) – AI for any task done beyond human ability.
RankBrain is currently classified as ANI. However, this can change very soon, as we currently live in times when breakthrough technology is revolutionizing the industry every day. It means that in order to maintain optimal competitiveness, we must be aware of all the technological advances and try to understand them in the best way possible.
How to optimize semantic search?
First of all, if you think you are optimizing for RankBrain, you should know that it is kind of useless because RankBrain is working when it encounters unclear or unknown queries. To optimize for those people, you will fight and lose.
Instead, what you should do is optimize the user experience. Yes, I think UX is becoming king pushing content out of its warm royal place. Of course, the content is still alive, but now users are the ones who tell Google whether any part of the content is good enough. This is a straight way (although it can be quite long) to guide the future.
And remember that it is wrong to think that search algorithms know more than we do. When Google tries to address search purposes, it uses tons of end-user data collected earlier to understand what users find relevant. So, it needs data to learn and data to guide once it has landed on your site.
4 important elements of semantic search
See what to do then to create high quality content that meets user intent and is completely search-crawler friendly.
- Create entities (Entity)
TL; DR: Semantic entities help search engines understand natural language and address search intent. Learn how to create them by applying structured data and becoming Knowledge Graph entities.
You know, we live in a new SEO world. Now that keyword research alone is not enough, you have to create semantic entities.
Semantic entities are people, places, or everything. The search appliance manages to decode natural language by understanding semantic entities, their characteristics and relationships.
In general, entities are at the heart of Google’s transition from the “strings to things” we saw with the introduction of the Knowledge Graph.
Knowledge graph of Knowledge Graph.
The Knowledge Graph (KG) is one of Google’s first steps to understanding how people see the world. According to the results of the Knowledge Graph, Google now has:
- A huge database of general information (capital, height and length, date of birth, etc.);
- Characteristics of each entity (for example, any location that is geographically located, may include smaller entities or be part of a larger entity, etc.).
What to do?
1) Apply structured data to the website.
To successfully create your own semantic entities, you must ensure that search engines will properly index and understand the connections between them. The best way to help search engines in this task is to apply structured data.
A universal structured data markup project, Schema.org, launched by Google, Bing and Yahoo to simplify and standardize how structured data is applied on the Web.
Go to the Schema page and choose the type of markup that best suits your entity (location, people, local businesses, events, audio, videos, images, objects, etc.). The good news is that you do not need to be a developer to add a schema to your site, you can do that with Structured Data Markup Helper.
When you finish adding the markup, you can test it with the help of Google’s Structured Data Testing Tool.
Moreover, the schema is also one of the prerequisites for your content to appear not only in popular organic listings but also in rich snippets, feature snippets and knowledge panels.
2) Become a Knowledge Graph entity.
The Knowledge Graph does not just refer to the tables to the right of the SERP. In fact, many SERP types are provided by the KG database. Once you occupy any such space, you will surely gain more visibility.
Before we figure out how to get there, check to see if there is already an entity for your business. Access the Knowledge Graph Search API, enter your brand or product name in the Query field, and click Execute:
Check the result. If there is no entity, the itemListElement array will be empty. If there is a KG entity for your brand, you might see something like this:
Or if you are not happy with the data currently displayed for your entity (it is automatically sorted which can lead to some ridiculous things), you can go to Wikidata. Use search to find a listing of your company, where you can edit details about your business, such as description, contact details, official website, etc.
Or you can suggest changes to your KG entity right from the KG dashboard in case you searched your business name and the dashboard was brought to you. Just click the Proposed edit option and then verify yourself to continue the changes:
If your business does not have a KG entity, try the steps below. We are not talking about immediate results, but these actions will bring you closer to becoming an entity:
- Use schema markup for organizations for your official website;
- Use a specific type of markup for your product (in the list of product categories);
- Create a Wikidata entry for your business and products. If you are not sure where to start, here is a guide for Wikidata;
- Create an article on Wikipedia. The problem is, Wikipedia is one of the main sources for the KG database. You can do it yourself, however, a good practice is to hire an experienced editor. Be sure to include a link to Wikidata in your section.
- Have your social media accounts verified by social networks.
- Pursuing search purpose.
TL; DR: The purpose of its search and effective solution is the number one task for Google right now. Learn how to optimize for search purposes by doing semantic keyword research.
It is clear that the day Google pulls all the resources imaginable to address the search purpose of queries, it is a high time to learn how to optimize it.
What must you do?
1) Create semantic keyword groups.
To perform better in semantic search, you have to stay away from old-fashioned targeting with individual keywords that should focus on broad topics instead. That means that now you do not have to create dozens of pages to cater to all your target keywords, you can create one or two pages that cover a specific topic in-depth and pursue the purpose. separate search.
No matter what new strategy you have, there is nothing you can do without keyword research. Rank Tracker can help you find a range of keyword ideas to form your semantic groups. Open the tool and go to Keyword Research.
2) Check out the keywords to be relevant.
It’s possible that Google uses the TF-IDF algorithm when it receives a query and goes through its index to find the best match.
TF-IDF is a formula for how often a keyword is found per page (TF: Term Frequency) and is expected to be found on an average site, based on a larger set of documents (IDF): Inverse Document Frequency – document frequency inverse.
Of course, it is not the only factor that determines the final outcome distribution. However, by analyzing the billions of pages and terms used on them, Google learns what they are related to, which words are synonyms and often appear in the same context. This data gives search engines an idea of what terms are likely to be in a given context for a given query.
In this condition, you should know the relevance of your target keywords when compared to related content.
I recommend that you go through your keyword list and choose the keywords or keyword groups that you intend to build your content around for the most important pages.
3) Develop content according to user intent.
You should clearly understand the purpose of bringing users to specific parts of your website. There are three types of common search purposes:
- Information – when people are looking for specific information and specific information and they are not buying;
Pro tip: As you pursue information queries, make sure you do not just focus on queries that can easily be Knowledge Graphs, which are questions that require completely real and decisive answers.
- Investigation – when people have processed information and tried to compare to find out what they really need;
- Transactions – when people come to a site with the intention of buying.
When you create content or reshape your content strategy, identify the common intent you try to pursue with your important pages. You need to do that to avoid unreasonable situations when your own pages get traffic from one page to another. It can happen when you pursue a few intentions with specific pages. Your pages should complement and amplify each other, not to compete for each other in ranking.
After you have dealt with the intentions you pursue, do a little research. Visit Google or any search engine and give them some queries with your target keywords. Research the first page (or the first few pages if you are patient enough) of the search list. Try to find the kind of intention for your keywords. Do not forget to find related searches at the end of the SERPs.
For example, if I google “red wine classification”, I mostly see different types of wines produced around the world, like merlot, zinfandel, carmenere, etc. However, the results are taken from Wikipedia as “Wine Classifications” and “Wine” and one result is just about the Grand Crus. In addition, related searches also suggest checking white grapes:
So, you see, we have the following intentions:
- classification of red wine,
- general information about wine,
- Complete wine classification and more information about a particular red wine.
Once you have discovered the intentions users often pursue with your keywords, check that these intentions coincide with the intent that you are trying to serve. If yes, do it well! You can detect some gaps that are not filled with the current SERPs and jump in with your content to supplement them.
If not, you must consider rewriting your content (or re-planning your content strategy) to better match your search intent or thinking of another page (or creating a new page) to target those keywords and the purpose.
4) Optimized for voice search.
It happens so that when we talk about semantic search, we also talk about voice search. Why? The problem is, when people search with their voice, they tend to solve their devices as if they were humans, using natural sound queries. They can be colloquial or ambiguous, and here comes RankBrain. So, voice search plays an active part in teaching AI to understand a natural language.
In addition, such voice queries are very intentional, since people often use voice search for some information they need immediately. For example, if I went back to my example with wine, when I asked for a voice search on wine, maybe I was standing in the supermarket and did not know what to choose.
The best way to optimize voice search is to think about all the questions that might be available to your potential visitors and then build your content around those questions.
You can think about these questions yourself or use some help:
- “People also ask” – these are a multitude of related questions and, again, a great way to borrow ideas from Google itself. Just google a question with your target keyword and, most probably, you will be delivered with this magical box. It is really magical because once you click on any question, you will see a series of new questions:
The best thing is that when you optimize for questions, you also optimize for featured snippets, because that is also the short answer to queries, but! You can see them above all the search results:
- Create topic groups.
TL; DR: People change their search behavior by using more specific queries, thus enabling search engines to learn how to understand the topic context behind a query. By building topic clusters, you provide context for search engines and introduce yourself as an expert on a specific topic. Learn how to rank higher by building topic clusters through managing your site structure and internal links.
Whether you have just started following a content strategy or continue pursuing it, you need to do it according to topic clusters. Why? The problem is, search engines have changed their algorithms (yes, by Hummingbird and RankBrain) to correspond to changes in user behavior. Instead of fragmented queries, users now prefer more specific queries with a range of keywords. And what’s more, they expect results to be delivered quickly and completely in line with their expectations.
As a result, these algorithms evolve in such a way that they can understand the topic context behind the purpose of search. In other words, search engines currently do not always need exact keywords to provide relevant results.
Therefore, many website owners now turn to clustered topic models. This model assumes that a column page is called the content center for a specific broad topic and that some pages link specific but relevant content back to the pillar page and to each other.
Such topology often signals to search engines that the pillar page is an authority on this topic and after scanning the entire topology, they will see that there is a semantic relationship between internal content of the pages. As a result, the site can be considered an expert on this specific topic.
To prove it, HubSpot launched a topic cluster test for a specific topic group in 2016. Their results show that the more they link, the higher the rankings are in search results.
What must you do?
1) Identify Pillar and Cluster content pages.
Think about the topics you want your business to compete on (and then base your keyword research strategies on those broad topics). Wide topic pages (aka Pillar or Pillar) should rank for short tail keywords, while Cluster Pages – for long tail.
A Pillar page should cover all important aspects of a really wide topic. For example, if one of your broad topics is technical SEO then you should have in-depth pages of usability, search, indexing, etc. Try to dedicate your Cluster pages to only one in-depth area mentioned on your Pillar, so that pages cannot steal traffic from each other.
2) Check your site structure.
If you already have a lot of relevant content, and you are not sure how it is connected (or connected at all), you can make good use of the Visualization feature in WebSite Auditor to see the site structure. and all the connections between the pages.
For the task of creating topic groups, I recommend drawing pages that relate to that topic with tags. This will allow you to instantly see how your potential pages are connected.
3) Manage your internal links.
Once you have figured out what your Pillar page is and decide on which pages should be part of a topic cluster, this is a good time to form a Topic Cluster.
- All Cluster pages must link back to the Pillar page, each page in your cluster must be linked at least once with the same keyword-targeted anchor text. In this way, search engines will know that it is part of a topic cluster.
- Linking from less content to more specific content will create sub topics and ensure a smooth PageRank flow.
- In those sub topics, link from high rankings to lower ranking pages to increase the relevance of these pages.
- Amplify your content.
TL; DR: Great content does not always show as you expect it to, so learn how to amplify its visibility with the help of visual media and social media.
Even if you are quite sure that your content is valuable and of great use, you can give it small visibility to users. When you remember, user experience can do magic when it comes to search engine rankings.
What must you do?
1) Do case studies and surveys.
It is not a daily task, but sometimes, you can put a lot of effort and time into researching some topics that are HOT or currently hot in your industry. You can run experiments and collect data to make some groundbreaking or justified conclusions. Or you can ask your friends and gurus some questions on some controversial topics.
Trust me, such content will be shared and referenced hundreds of times and will appear in search results for years.
>> Learn how to deploy Viral Content
2) Add useful visual information.
And I am not talking about memes. Although they can certainly amplify what you are trying to say or what you feel about any point. I am talking about infographics, diagrams, workflows, training videos, and more. People prefer to actively refer to such information in their own content, thus providing you with natural backlinks.
Moreover, search engines are getting pretty good at understanding the images, audio, and videos.
3) Available on Social media.
Experiment with social media platforms and choose the best ones to amplify your brand. Share your content there to reach a wider audience. Start discussions to find out what your current or potential users are missing or what they are happy or sad about. In addition, social media is a great source of user-generated content, often endorsements from colleagues.
4) Get customer reviews.
When people write reviews, they use natural language and describe the advantages and disadvantages of a specific product / location / service. Such content helps Google analyze and return this business in search results for queries such as “the best software for checking rankings at an affordable price”, “ezcloud is hotel management software good, “etc.
So, whenever you find customers happy, try to encourage them to leave a review. Positive reviews will show you on Local Map and Local Knowledge Graph. Just make sure you have a complete Google My Business listing.
The danger of AI
While we are happy to enjoy the smart inner workings of search engines that give us the information we need and even more than it requires, there are some AI trends that both users and users alike. Using SEO is not happy about:
- Pushing out the organic list.
Obviously, we said goodbye to only ten blue links (10 results in the search page) when searching for some common things. Screen space is currently available for all types of results from Google’s database: Knowledge Graph (usually provided by Wikipedia), Answer box, Image carousel, YouTube Carousel, etc.
Moz ran several experiments in the process where they found that SERPs with 10 blue bonds received about 79% organic CTR. When the Knowledge Graph is distributed, the organic CTR drops to 25%. No comment.
On mobile, you won’t even see organic results unless you do some scrolling:
- No search results.
I mentioned this strange trend at the beginning of this post. When you start typing your query in the search bar, you can immediately see the answer if your query is completely realistic:
It means you will not even go to search results if your query is satisfied!
Another concern is the recent experiment with zero search results: people googled for some practical information, for example, the current time and they see this:
Yes, answers to their questions, free space and buttons show all results.
Please do not worry, the test is closed, now. And although such SERPs are only distributed for local time display, currency conversion and calculations, if it becomes a reality, it can also be extended through other types of results. For Google, it can deliver results faster, thus being more efficient. For SEO-ers – is a disaster.
- The results are overly personalized.
I think you know that two different users may see different SERPs for the same query. Typically, this happens when users are in different locations and they search for results based on locality. Or when they search from mobile or desktop.
Google claims that they had a light on personalization focusing solely on location and language. However, if you experiment a little, you will find that if you log in to your account, many factors influence your search results, like browsing history, browsers, social media profiles, etc.
In such situations, your results are really biased. In most cases, you yourself want them to be biased because they include your personal intent better. However, when you want to get rid of this biased reality, you really need to take some precautions, which will not make you feel completely at ease.
From the perspective of search engines, it is easy to understand why they pursue the development of semantic search. It means more data, less spam and black hat techniques, deeper understanding of natural language and search purposes. All these results in providing the best search experience possible.
However, it can now be said that the search trend is changing, but gradually, not exponentially. You still have to follow the same optimization methods, like keyword research, link building, UX enhancement, etc. The only difference is that you need to do it with the idea that you optimize for both users and search engines and for search engines that can now understand the meaning of a query without the main keyword.
Bizzvn wishes you to apply the knowledge from this article to optimize good Content with Semantic Search!