Table of contents
- Keyword analytics, is it up-to-date anymore?
- Entity analytics – a different approach
- How entity analytics and SEO entity performance work
- What if our system has many entity pages?
Keyword Analytics, Is It Up-To-Date Anymore?
Modern search engines and retrieval systems have shifted from keyword-based search to entity-based search. Therefore, as a result many information needs today involve searching for entities and understanding the relationships between them in order to obtain the right answers for searchers’ needs. In simple terms, this means that instead of relying only on term-based searching and term-oriented ranking, we and search engines will exploit the structure of the knowledge graph. That is how entity analytics is born.
In this article, we will focus on understanding how entity analytics is beneficial for developing modern SEO strategies and how it is different compared to previous keyword-based approaches that cannot be used anymore to establish a competitive SEO analysis and SEO entity performance in the modern world.
The casual process of performing keyword research and analysis is to gather keyword information from multiple sources like Google Search Console, Ahrefs, communication strategies or sales data and then obtain the keyword ranking information or keyword volume data points. Then, you would use this information to filter out preliminary queries that fit user needs, calculate their potential for topical authority and embed them into your final content plans.
To measure topical authority for a topic (eg “e-commerce”), we could measure how much (estimated) traffic a site gets from its underlying keywords. Example: if a site gets around 15% of available traffic for a topic and it is the highest share, then this site must be the most authoritative for that topic. Kevin Indig wrote an awesome article covering measuring topical authoritativeness based on defined keywords for a certain niche.
Entity Analytics – A Different Approach
However, with entities in mind, the process is a bit different. We can take two individual approaches for SEO entity performance and entity analytics.
The first one is focused on working with entities in the following way:
- You can get the top search queries your website or webpage is ranking for through keyword extraction techniques (try KeyBERT);
- In the next step, you can extract the most recurring entities from the pages behind these search queries by using Google NLP API or your own pretrained model for entity extraction (we use ours that is trained on DBPedia data);
- In the third step, you cluster these selected entities by using:
- graph embeddings, a well known topic in the computer science world;
- SBERT using the entity description from DBpedia;
- wembedder or BigGraph.
- Use the previously obtained information to analyze the topic you are interested in.
To bring more clarity to this approach, we need to explain what graph embeddings, SBERT and wembedders mean. In simple terms, here is what they do:
- Graph embedding is the process of transforming nodes, edges and their features into a vector space while preserving all these properties in a graph-like structure and information.
- SBERT or BERT for sentence similarity is a so-called twin network that processes two sentences in a similar way simultaneously. It is often used in tasks where the model performs content clustering and classification.
- Wembedders are embeddings trained on Wikidata, usually used to perform nearest neighbor search.
The second approach that you can use to apply entity analytics to your SEO strategy is using the SEO Add-on for Google Sheets™ by WordLift that we are also using for our clients’ work and strategy planning. The plugin does the following:
- Analyses your content in detail and performs entity resolution. “Entity resolution is the process that resolves entities and detects relationships. The pipelines perform entity resolution as they process incoming identity records in three phases: recognize, resolve, and relate.”;
- Once the entities are detected, they are connected to their respective knowledge-base definitions to be properly disambiguated;
- A knowledge graph is built.
Start using SEO Add-on for Google Sheets™
Perform your semantic keyword research in 7 simple clicks.
How Entity Analytics And SEO Entity Performance Work
This is completely different to keyword analysis, because a separate page for every entity is created accordingly. So imagine that you have an e-commerce web page or you are a publisher like The Next Web and you want to know which pages or news articles are the most popular but on an entity level. Creating those separate entities will allow you to inspect analytics for every separate entity. So instead of analyzing page performance rich of unstructured data like text, you will be able to do analysis on a micro level, inspecting every entity individually.
Imagine what this could mean for you if you are an e-commerce platform and have multiple cars featured there. We are sure that you would like to know what your audience characteristics are for Audi or Tesla models for example. Our entity-based system establishes exactly that: micro-level analytics focused on entities through entity building. Every page reflects a separate entity, so every data point obtained from Adobe Analytics or Google Analytics will point you to the data points for a single entity instead of analyzing keywords for non-disambiguated data (text web page).
What If Our System Has Many Entity Pages?
When dealing with multiple entity pages, you can perform entity clustering to obtain only the most interesting and dominant entities and filter out the rest. In fact, this is a recommended approach because over time, the entity corpus is going to grow more and more and some data cleaning and crawl budget optimization will be needed.
Entity analytics is a powerful technique when used in the right way and with the right tools. It is just a matter of time when it will get adopted by the wider masses but if you like to remain competitive in your niche, you will definitely need to try it out.
Must Read Content
The Power of Product Knowledge Graph for E-commerce
Dive deep into the power of data for e-commerce
Why Do We Need Knowledge Graphs?
Learn what a knowledge graph brings to SEO with Teodora Petkova
Generative AI for SEO: An Overview
Use videos to increase traffic to your websites
SEO Automation in 2023
Improve the SEO of your website through Artificial Intelligence
Touch your SEO: Introducing Physical SEO
Connect a physical product to the ecosystem of data on the web