Query Semantics SEO Case Study: Convince Search Engine to Change Meaning of a Query - Two Websites

Key Takeaways

hello my name is correctober and I am the owner and founder of holistic SEO and digital in this video I will explain and demonstrate two different websites with the semantic SEO and at the same time I will try to explain the concept of the query semantics and a little bit different its difference from the lexical semantics as well in some of the previous videos I actually mentioned that I will launch a semantic SEO course still I am trying to actually arrange some conditions while waiting the world to be ready and before this launch I will publish many different SEO case studies if you just read and watch these videos I am sure that you will be successful in the semantic SEO course as well and you will get a certificate because in semantics it's not like technical SEO as I say there is no determined or predetermined documentation it doesn't use a universal language linguistics and natural language processing it is highly complicated area without a single Center and especially when it comes to understanding a semantic search engine is way much harder so with that said if you want to learn when it is published or when it will be published just subscribe to these holistic as your newsletter and then actually when I publish these case studies or the when I launch the website for giving the semantics your course to you you will be you will get the news besides that probably if you watch these things I guess I will be all this article will be already published but the thing is if you check this project and these one they're actually coming from the same environment this one is a product search engine in other words it is a little too unique because uh how can I tell it it is too unique because it is a product search engine but beyond that because it doesn't sell something online it just tells users to search this thing online and then go to a specific store and just try it physically so they try they have to track all of the physical storage or they have to understand the stock information so that they can tell you that yes if you like these shoes you need to go to these store from this city in this area and it will be open on until to the death point so this is this kinds of uh search engine in other words it is really unique search engine so as an SEO it is also a good feeling that you are doing actual search engine optimization for a search engine design which is a unique new search engine type and I like actually doing that and I can show you the results here too once I published actually these to read by telling that there is no discipline in the client because the client didn't even realize that they launched the project without perfect conditions and anyway this was a quick launch a quick increase as you see suddenly and this is the initial coverage report I can publish these things easily because it won't harm the project because most of this data is already outdated and I'm sure that you won't be able to create same type of website and here too these are the updated results nearly updated results essay from the hrefs as well when it comes to this one this is also important too because this project is a B2B e-commerce but to be able to be successful on the search engine optimization sometimes you have to augment the queries and you need to connect them to your own context that's why it is important to understand the query semantics because this project is for gift selling but not for the users or consumers directly it is actually the four uh B2B selling with a really big amount of materials or the product counts so the problem here is that no one searches it properly even if they search it they just search maybe 56 queries with a certain type of phrases phrase combinations but when it comes to Turkey since this is a touristic country you have we have many cities ministeristic environment and all these gifts are actually they actually have touristic team so to be able to connect these cities to the tourism than to the touristic gifts and then to the B2B selling you need to understand the importance of the b2c there and also importance of the importance of the objects that will appear in the in the gift objects as well so in this case you can't directly go there and create a B2B website you need to change or balance things from the search engine point of view so that they can actually see you as a candidate there so in other words if someone searches for taking a single gift for from let's say Antalya with a specific type of team with a specific type of object or something it means that actually you have to con you have to be connected to that user so that you can also be connected to the users that actually want to buy B2B products or let's say purchase maybe even two millions of entirely related gift products so the another problem here is that even if the a person actually is or let's say has B2B business still they don't use actually these type of queries because diver the consumers and also be to business owners they usually use the same queries and search engine can't always predict who is who or why the person actually searches it that's why they have to look at the user demand and predicted user profiles so in this case of course user clustering and query clustering and matching queries to the users these are also important subjects but we can talk about audience clustering maybe later in this project to be able to rank for B2B queries I had to create actually a direct consumer website first and then I needed to to Vick the selling funnels or let's say sale funnels and also some buttons even or pricing methodologies or other things because if you go to the product pages of this website you will realize that you can even order 100 thousands of props or items suddenly or you can just buy one too so it is actually selling to the consumers but they are not the main target there the problem as I say is query semantics because when it comes to B2B gift related searches it's two less the search demand is two less and they are so blurry there is no specific type of Border or search Behavior differences between these two different distinctive audiences so it means that we need uncertain inference and then we will need to cover all these contexts that will come from these queries this is the result from the ashrefs and this is that from the sem rush and this is from the Google search console data as you know I actually published these things for balancing the page Rank by explaining cost of retrievable Link value proposition and information responsiveness related things and the cost of retrievable also will be explained later too these are some of the initial results I can tell that project is really slow because both of the projects have highly technical SEO problems and there is not enough level of enough level of let's say [Music] outer Force let's say in this case and this is gerald.com it is around here as I I can tell it actually if you go to the this area you will realize that it has actually zero domain rating from According to the ashtrays but these are the initial metrics there are some problematic things with these websites I actually explained them into these aggregated and curated background of subject websites with SEO problems you can read these areas to understand why they were slow why there there were some other things but I can tell that with these conditions without any kinds of Link or page rank I believe these things are creating good results and there was a strong ontology and taxonomy to create a proper website tree and especially when it comes to girard.com the main problem was actually URL structured because we can't use coordinates in the URLs so that we shall not create millions of endless URLs to the search engine in a without a meaning if you do that as Bing says here user demand is important thing to understand whether there is actually a kinds of spam or not if you create too many pages without any kinds of user demand search engine might assume that it is actually spammy bed page is that can be generated by a machine when it comes to Google if there is no user demand or proper user demand they assume that it is an unpopular topic but still it doesn't mean that it might not it might not be posted as pump 2 so it means actually you are not entirely in a safe Zone besides that there are some other things of course I need to mention so let's read some questions together then we can try to answer them as well what is query semantics why is it necessary how do search engines use Query semantics why is query semantics important for B2B industry why is search demand matter for anti-spam algorithms why do search engines need to need to variate meanings what is the difference between lexical and query summon things and how to evaluate query semantics for some rare SEO projects so query semantics that means that actually the meaning of the query changes according to the query language the query language concept comes from actually Crown field experiments ground field experiments maybe I can find them because in crownfield experiments the users this is one of the first information retrieval researches and it is highly important actually even for today's information retrieval systems and in the crown field yeah here the query language has been used and it is the first actually acceleration of the query language basically how you search is your query language and in the query language we don't use meaningful sentences in as in the natural language we you just write certain phrases side by side and we expect machine to fill the blank areas which actually brings us to this concept which is uncertain inference because every according to uncertain inference we always need an old space I won't go or dive in here because I have explained it I believe before I believe I explained this thing already so if you just try to read a couple of Bing blog posts you can see that actually they are useful as much as Google's own blog posts but based the query semantics means that every query changes its context and the meaning according to the every signal and denote denotational note or structural synthetic and also semantic unit or lexical unit even their orders can affect actual relevance and search engines have to actually generate questions sentences or inferrings inferences from from your queries or queries in the query languages it is necessary because search engines have to understand the possible contacts once they understand these contexts they have to actually prioritize them and then they have to match these contexts that come from queries to the contexts from the documents let's save it five percent this is the context of the query but in this document let's say 30 percent of the document is about this specific context it means that actually if this section is too early in your document you will you will need to lose the rankings you have to parse the query as the search engine purchase then you have to match every contextual section or unit according to that as well and with that said there will be many other things in this area to be honest for example here when you look at to do this research actually it comes from Google let's say not entirely Google but it is a mutual project actually in the Google research publication section you can find it and this section actually is not directly about the search engines but it is about the search and they explain what rdf is and why how to use it for actually query language and query by example here actually they try to take the examples then they try to connect all these samples to the queries and then they try to create a taxonomy between these things so that they can understand what part of toxonomy is a better match for your query you can actually match these things with my example too when it comes to the resource the encryption framework you already know it from the structured data we usually use actual different formats but you can also check this and there are some other examples here in this section too if you just check actually W3 they will explain you what a query is why it is important for semantic web you can use actual W3 Consortium as a information Source or basic information Source it will help you you can understand what a triple or what a pattern is here and with that said this is the Google research publication that I just showed you there and at the same time when it comes to the semantic queries this is also important in a way because sometimes you can use these knowledge spaces some of these knowledge spaces are actually for creating applications but in this case if you want to create let's say you want to understand the search engines it doesn't mean that you should create one but at least you can try to have an empath with the search engine Engineers that's why these knowledge pages are actually important before coming to that area actually this is also an important blog post again from the Google to explain how AI helped the Bing to have better query understanding and there are some samples here for example when you search for Windows update reset tool before and after they try to show actual how the results change according to the semantic semantic relevance and and they also explain how these things actually can be used in other cognitive cognitive search as well and here actually they also explain how you can create your own semantic search engine and how you can weight the different types of places on the web pages or how you can actually parse the queries and Etc is up to your own configuration but besides that from actually Microsoft pink This is highly important another post to understand the query semantics because when it comes to the query semantics you have to understand that a word will definitely change its meaning from in in a dictionary if let's say something is synonym with something it doesn't mean that it will be the same in the query language sometimes it might sometimes it might not be sometimes it might be in that way if a third another concept appears them appears there with together with them so here when you search for how long does it cancel the last chance can't can't diet soda soft drinks also come there unopened room temperature pop also kind of comes there carbonate drinks come there so according to the Microsoft being canned soda and these things are synonyms to each other or interchangeable phrases so in this case you have to understand that actually these relevances actually also come from word to vac models and glovie type of things and they try to understand what is the canonical query according to the query format and at the same time Concepts in the query there are other things here to explain the character embeddings if you want to understand the prominence of the character embedding I would suggest you to check these two parts it will be helpful to understand how it increases the Precision even further so I believe these sections explain them but doesn't end there because enriching query summonings for code search the code search it goes richer every year and now even actually Google for Transformers or Google language models the code generation is another NLP task or language model quality task or Benchmark as regular natural text Generations in other words when you say something the machine needs to write the code in a certain programming language so this is about that and how we can actually match or transfer human language the code language or programming language this is about that and query someone thinks here it is needed because if someone searches or if someone actually tries to generate code in in as in the query semantics it will be harder and here they try to understand the semantic gap between them so I believe I have explained the query semantics and why it is necessary because search engines will need to associate cluster and also predict the every meaning every context of all these possible inferences that come from the queries then they will need to distribute these possibilities to the audiences according to the user demand they will need to match the documents according to that and sometimes document popularity might affect the query context if there is not enough level of user Behavior or user data so that because search engine will assume that if if the query search demand increases then this type of document suddenly appears it means that these documents are directly for these type of queries as well so in this case when I create my own content briefs or semantic content networks I have to remember this type of possibilities so that I can put my source in front of others especially for these type of seasonal SEO events because they will show who is more quality and who is more needed on the serp or on the web and why the search engines need to variate the meaning of the phrases in the query because first of all in the query every word has multiple other types of meanings and a search engine needs to always actually change the meaning of these things because the context always changes so a word might have nine different contexts at but when you put another word before or after it is Con there might appear some context constraints and according to these context cons range the search engine will need to start actually filter these documents but as you know serving all these documents faster or faster and faster is more and more important in this case a search engine might start to realize that instead of filtering these things out already we can actually register every document to do a different indexing chart or index petition or partition so in this case we can actually store them faster it means that actually documents might change their places in the indexes or indices and this might happen according to the query context so and that's why actually having a contextual consolidation is really really important and the document here the searching with the context here it actually explains many different types of possibilities and explanations here I would suggest you to read this document or this paper to be honest because especially ranked biasing or query writing query ranked by singer besides that how this iterative search processes how they can actually change ranking with every iteration it is also important for understanding the cost of these things as well and also it is important because sometimes search engines might not choose every word in the query to rank the results sometimes even if you delete a word from the query results don't change right so it means that actually it is not a contextual word or it is not a necessary word it doesn't have a weight in the query here they Explain how other queries can be generated according to the context and difference between lexical and query summon things is that the in the lexical semantics come from nature language query semantics come from search language or search query language in this case as I say before to selling and buying they might be antonyms in the dictionary but they are synonyms in the in the query semantics when you when it comes to using all these information actually for SEO or search engine optimization you will you will need to use actually this type of understandings or these type of approaches to create or design your website because you can't directly create a website for only b2c because b2c or B2B you will need to calculate all these possibilities and you will need to augment your top your topical map if Google infers a query in a different way and if they augment the meaning of the query it means that you will need to augment your topical map too you might even need to change some certain sections of your business model or business identity or Source context because as I say in in some of the previous videos and also in this information responsiveness related case 32 these things matter because in helpful content update announcement to Google site that actually who you are and why do you explain these things so your identity and topic should be relevant to each other if you change your if you change the queries that you want to rank it should be connected to your Source or your identity and to be able to provide that you need to position your identity or the brand according to that as well so I believe this was a quick video according to the others I hope you enjoyed it I have I try to explain many things here by going over these things I will try to put more videos soon and as I say probably this document has already been published or if it is not I will publish it I might publish it as a book too because it is already too long and I don't know whether people will read it or not but usually I write for myself I can't even tell that I take some of these videos actually for training some some teammates in a faster way so I also share these things so that other people also can benefit them so thank you for listening to me and I promise you that I will publish and launch the semantic SEO course but when 2 000 people or three or four thousand people start to ask me questions it will be so hard for me so let me prepare you please watch these videos and read this as your case studies and take care of yourself bye