Author: Michel Fortin
Traditionally, Google would rank your website according to keywords in your content and pages. The number of keywords and the number of times those keywords were mentioned helped to determine the relevance of the page (and they still do, albeit to a lesser extent). However, due to overwhelming spam by some website owners who stuffed their content with keywords, Google has since improved by moving away from keywords and towards something called “entities.”
Entities are ideas. They’re concepts that represent places, people, things, events, etc. They also have various attributes and connections with other entities that help search engines understand the context in which they’re used. For example, if I mention the word “robot,” you might think of a physical machine that performs a certain function. But depending on the context, a robot can also be a piece of software, such as a search engine robot that crawls your website to include it in its database.
How do you know which one is which?
This is where context comes into play. If the word robot is used in a sentence, paragraph, or page in which search engines are also mentioned (or if the word robot is used to describe a particular search engine crawler), then you know it’s the software kind—such as Googlebot.
To serve the most relevant results, Google is continuously learning from your brand and about it, just as it’s learning from your users. It’s trying to determine if you’re the right fit for what users are looking for. To add more context to a piece of content, Google tries to understand the entity through its relationships with other entities. So, it looks for connections between them and associates different entities together in something called a knowledge graph.
Those connections are vital in the world of semantic search. It’s more than helping Google make those connections. It’s also about increasing Google’s confidence in them—that the connections are correct and make sense with respect to the entity it’s related to. The goal, therefore, is to disambiguate your content and make it easier for Google to learn from you. The more it learns, the more confident it becomes and the higher the chances are that your content shows on relevant search results.
This is where semantic SEO comes into play.
Disambiguation tips to improve semantic SEO
While it’s important to provide great content, your goal should be to provide great clarity, too. “Create quality content” may be Google’s mantra, but once you do, your goal should be to create as much clarity as possible with your great content.
In other words, help Google identify the concepts in your content (i.e., entities, attributes, and relationships), connect those concepts together, and increase its confidence in those connections.
Semantic SEO is about disambiguating your content as much as possible. Ambiguity can be your biggest stumbling block when trying to rank well as it creates guesswork for Google—and Google is a machine, so it doesn’t make guesses unless it’s confident it’s making the right ones. Disambiguation helps reduce that guesswork and makes things clearer, which can significantly increase Google’s confidence in your content’s topical relevance.
To help you, here are disambiguation tips that will improve both clarity and confidence.
01. Write content with clarity
First and foremost, you should always write clearly and concisely, using plain language that’s easy to understand. You may have heard the saying that you should “dumb down” your writing—I don’t like that statement because it’s not about making your content less intelligent. It’s about making it easier to read and accessible to the widest possible audience.
Quite simply, avoid cryptic language or trying to be clever. If you’re unsure, run your content through readability checking tools to evaluate the level or grade it is currently written in. These tools use various formulas (such as Flesch-Kincaid and others) to rate the level of education someone needs to read your content. Or, use software like Grammarly or ProWritingAid, which can also check your content’s readability.
Google has introduced natural language processing (NLP) into its algorithm, which means that it can process and understand human language. Simple language is therefore good for both people and Google. If your content forces users to do mental contortions to understand it, even if only slightly, it will make things even more difficult for Google and potentially prevent it from making those important connections you want it to make.
In short, if people have a hard time understanding what your content is trying to say, Google will have a much harder time understanding it.
02. Enhance pages with structured data
Structured data (schema) can help Google make better connections with the entities it finds. Schema is metadata—the data beyond (or behind, if you will) the data—that can help Google understand what your content is about. Look at it this way: it’s like a narrator in a movie explaining what is going on that may be unsaid or unclear. Schema is like giving your data a narrator or a voice, if you will, that explains what the content is about.
For example, let’s say you’re a baker and you sell apple pies. If the word “apple” is mentioned in your content, and it’s about the fruit and not the electronics brand, then you could add schema about your local business (e.g., “Michel’s Bakery”) or your products (e.g., “apple pies”) to the code of the page. There are all sorts of schema that you can add, from business information to recipes to FAQs. You should add applicable schema to your site, and platforms like Wix offer tools and guides that can help with this process.
03. Structure text for readability
Adding schema doesn’t mean you can ignore your content. Schema is not a free pass or a magic bullet. It’s important to make sure your content is well-structured so that it’s easier to read and understand. After all, Google must find entities within your content and understand the relationships between them, so avoid making your pages look like disorganized blobs of text. What’s more, poorly structured content results in a confusing user experience, which can ultimately result in lost conversions.
Grouping topics together (and adding headings before each group and throughout the page) can help distinguish important sections and introduce the topics that follow. For longer content, adding a table of contents at the top, and including links that jump to the major sections, may help users find relevant information faster. More importantly, headings and internal links provide context when they’re topics (i.e., entities) as well, such as linking “bakery” with “apples” mentioned in the previous example.
04. Delineate sections with markers
Just like headings and subheadings mentioned above, visuals are useful because they help support and contextualize the surrounding content, too. They’re also useful in helping to identify, clarify, and amplify concepts and their relationships—and possibly adding to the knowledge graph, too. Similar to headings, visuals serve as markers that provide cues for both readers and crawlers, including Google.
For example, if the word “apple” is mentioned alongside a photo of an apple pie, then the connection is made obvious. More importantly, since people tend to skim, scan, and scroll, cues help users move from scanning to reading. By the same token, they also create landmarks that are important for accessibility (e.g., visually impaired users with screen reading tools). Semantic elements help structure the content in a way that both humans and machines prefer—these can include:
Headings (e.g., H1, H2, H3, etc.),
Lists (e.g., bullets or numbers),
Emphasis (e.g., bolds, italics, highlights, etc.),
Visuals (e.g., images, photos, graphics, icons, etc.), and
Descriptions (e.g., titles, captions, alternate texts, etc).
However, just as it is no longer necessary (or effective) to stuff your content with keywords, it’s important not to go overboard here, too. As the saying goes, “When you emphasize everything, you emphasize nothing.” Plus, too many markers will take away from the clarity it’s supposed to help create. Use them judiciously, moderately, and strategically.
05. Repeat ideas for comprehension
With lexical search, it used to make sense to include synonyms and keyword variations in order to capture and rank for all the possible versions. But with semantic search and the help of natural language processing, there’s no need to explicitly use variations or to repeat them needlessly. Google can understand how pages relate to a certain topic, often without containing the exact keywords or keyword variations.
However, you may express some ideas in a completely new or unique context. Users may understand what you mean, but it might still be unclear to Google. More importantly, if there’s a possibility that your users can misunderstand or misinterpret what you mean, Google will likely fare even worse. Therefore, repeat new ideas using variations or synonyms, or express them in different ways. This will help remove ambiguity and increase Google’s confidence in your content’s relevance to the user’s intent.
Take the term “seal.” A seal can be an animal, an indicia, an envelope flap, a waxed imprint, a certificate, a box closure, a pipe ring, the pop singer, and so on. Google knows all of these meanings. But when a search comes up for one of these, how does Google know which one to show? If we narrow it down a bit more (and to borrow an excellent example from Dave Amberland), let’s say I use the term “trained seal.” This narrows things down a bit more, but some people may think either one of these two:
To a non-American, the military “trained seal” may not be immediately familiar. Nevertheless, if I use repetition but with variations, like the words “sea lion,” “circus seal,” or “phocid” (the scientific term for seal being from the “phocidae” family), I’m clearly referring to the animal, not the soldier.
06. Optimize entities for recognition
As mentioned above, while Google is sophisticated enough to recognize most entities, it may be missing some. If you create content about unknown terminology or completely new entities, Google may misinterpret or skip them altogether. So, run your content through an entity extraction and recognition tool to see what entities it finds, as well as which ones are ignored or mislabeled. You can then try to reword your content to make it easier for robots to understand (and, potentially users as well).
You can use a named entity recognition (NER) program, which can extract, identify, and segment entities in your text, and then categorize them under various predefined classes. Note that these may vary depending on the tool. In addition to Google’s NLP online demo mentioned earlier, there are plenty of entity extraction tools such as demos of entity search engines and open NLP software you can use.
07. Contextualize with internal links
In SEO, one of the most important and recommended on-page optimization practices is internal linking. Use it to disambiguate entities, too. Link topics and subtopics within content to other pages they relate to. Contextual links help Google understand what the linked page is about and associates topical relevance back to the origin page. Think of Wikipedia, which does a masterful job of this.
The anchor text (i.e., the text that links to another page) contains words or phrases that help to disambiguate the entity and its category by relating it to a page that’s on-topic. For example, if you mention the word “cats” on a page without any context, and you have another page that’s all about pets, you can link that word to that page, which makes it clear that you’re talking about the feline persuasion (as opposed to the Broadway musical, for example).
08. Cover topics fully and frequently
According to Google’s Quality Raters Guidelines (QRG), it’s important to meet the needs of searchers and to provide them with quality content. One aspect Google looks for and asks their human quality raters to measure is “comprehensiveness.” For example, does the page provide an adequate amount of information on the topic? With semantic SEO, thorough content also helps to improve the clarity around the topic.
The content must contain sufficient information to help clarify what is being discussed. But covering all the bases doesn’t necessarily mean writing an entire encyclopedia entry. Google has released an algorithm update whereby it penalizes websites that contain unhelpful, SEO-driven content instead of user-first content. This applies to overly lengthy content as well. Just think of what your user needs to know along their journey or what questions they may have about the topic. Plus, it doesn’t have to all be on a single page—use contextual links to connect subtopics together where appropriate.
Similarly, publishing on-topic content frequently can reinforce topical relevance. But staying on-topic is just as important. Subtopics and related topics are fine, but refrain from including unrelated ones on the same page. Unless they’re related somehow and fall under a larger umbrella topic, keep the main topic in focus. Otherwise, needless variety creates ambiguity and may dilute the page’s topical relevance.
09. Group topics into clusters
When multiple pages cover vastly different topics, clustering topically similar content together is helpful for usability. It makes content easier to locate and discover. It also helps users self-orient and navigate your site (such as with breadcrumb menus, tables of linked content, and URL structures). Topic clusters also aid disambiguation by grouping and contextualizing the content from related pages. You can use folders, categories, tags, labels, collections, and so on. URLs may look something like this:
This is not about adding keywords in URLs (which is highly unlikely to substantially improve your SEO). Neither is it about making the URLs longer just for the sake of making them longer. It’s simply about contextualizing sections of your website and grouping topically relevant pages together. It improves the user experience and helps both users and search engines understand a section’s theme or function.
However, even if a page falls under multiple categories, it can only have one URL. (If it exists under multiple categories, use canonical tags to prevent duplicate content issues.) But, it can be the primary category or a larger one, such as an umbrella category that logically covers a series of subcategories. Even with single-category folders, connect topically related content together using contextual links.
10. Answer questions using knowledge graphs
Help Google make connections by adding to an existing knowledge graph. When you’re writing about a well-known topic, a good strategy is to answer questions users may still have around that topic or clarifying any information that its associated knowledge graph already contains. This way, you’re adding new connections to an existing knowledge graph, which should be easier than trying to convince Google to create a new knowledge graph entity from scratch.
For example, instead of conducting keyword research, analyze the questions people are asking and answer them. Search Google and look at the “People Also Asked” section halfway down the page. If the answers are part of an existing knowledge graph, then take it one step further by linking your answers to its knowledge panel (i.e., a panel that appears in search showing information related to a knowledge graph). Even Google Business Profiles (GBP) can be used as a means of answering questions and connecting them together, such as with the products and services listed in your GBP.
When answering questions your market is asking, answering them in your content, and connecting them to your product or service, make sure to add schema markup. It may include Q&A schema, FAQ schema, or eCommerce schema (i.e., product or service schema)—including schema related to media you’ve embedded in your content, such as visuals, videos, photos, etc.—to make it easier for search engines to understand how to make those connections.
“What” is never as important as “why”
Previously, it made sense to focus on what people are searching for. But now, it makes more sense to focus on why they’re searching for it. Think of it this way: your content is trying to answer the questions that your user is asking. But without any context, it’s like jumping in and interrupting your user’s question midway through, and giving them the wrong answer—or the right answer but for the wrong reasons.
SEO aims to help search engines better read your content. But semantic SEO aims to help them better understand it. And by understanding the context behind your content and the intent behind your user’s query, search engines can create more meaningful matches between both. The more meaningful the match is, the more meaningful the traffic it generates will be.
Michel Fortin is a marketing advisor, author, speaker, and the VP of Digital Marketing at Musora Media, the company behind Drumeo. For nearly 30 years, he has worked with clients from around the globe to improve their visibility, build their authority, and grow their business.