Craft fact-based content that maps to the search engine’s understanding of entities
Jamie says: “Our reality changed so much over the last few years that we were forced to order groceries online and use internet search to find out whether it was safe to go outside.
As consumers, we have collectively embraced a digital-first mentality. In 2023, we SEOs must keep step with Google as it’s pressed to combat myths and disinformation in SERPs. We can do this by crafting fact-based content that maps to the search engine’s understanding of entities.”
Is Google getting better at combating myths and disinformation?
“Yes, it is - and it’s had to be more vigilant in this area. Today, if a hot news trend emerges but there are no factual or authoritative sources speaking on it, you’ll get a bit of a notice that indicates there aren’t really any good answers. Google has had to acknowledge that, even as a search engine that wants to give answers for everything, there are times when those data voices can be weaponised and shouldn’t give answers for everything.”
How would you summarise the method that Google currently uses to understand what an entity is?
“The easiest way to understand it is using an analogy from The Hitchhiker’s Guide to the Galaxy. It was a book that the main character, Arthur, carried around with him. It contains information on everything that exists in every person, place, colour, thing, concept, event, etc., all in one handy spot. Google has its own version of The Hitchhiker’s Guide to the Galaxy which is known as Wikidata.”
In the eyes of Google, is getting your entity into Wikidata the fastest way to get your brand established as something they understand?
“Absolutely; it’s the cornerstone foundation. It’s important to remember that, technically, entities are in two places. If you’re just getting your business started, you’re probably going to begin as an entity in Maps in Google Business Profile because you have goods to offer. This would be your first cornerstone. The larger scale suggests we look at the concepts that we try to rank for in our head-level keywords. These are our entities.”
Are you thinking of schema? Are there different ways of marking up your business and different facets of your being as an entity?
“Absolutely. With structured data, if you were to go ahead and open up Wikidata on a specific subject, you would see things you recognized - perhaps from books, structured data markup, the year the book was published, the number of volumes it’s had, etc. Those pieces are all copied over. When we talk about this great inbound Web 3.0, one of the cornerstones is an interlinked data structure. In the same way that we write up schema for Google to understand what the web page is about, they have their own schema to understand what our reality is about.”
You mentioned misinformation and disinformation at the beginning. Are many websites marking up their data incorrectly and/or is Google having to fight this as well?
“Incorrect implementation of markup would not be the same as accidentally sharing misinformation. You read a headline, you write about it, and you don’t fact-check it. Despite there being schemas in place to understand the world better, just like any other SEO elements, those are often abused.”
You recommend the creation of fact-based content. What is fact-based content and how do you go about creating it?
“Let’s say, for example, we were to have a well-informed conversation about Star Wars. We’re probably going to casually mention Wookiees, Siths, Ewoks, and Jedis. This is all part of being well-informed on the topic. There’s semantic interlinking, the same way that in Schema.org information you could mark up a book with characters for Wookiees, Siths, Ewoks, etc. - they’d have the entity mappings of the same things. They know how the pieces interconnect and create factual content it’s going to relate to. Are you covering the facets and understanding of an entity expected to be considered an authority on it?”
If you’re talking about Star Wars but you’re not a film website, would Google not assume that you’re going to be talking about the film Star Wars?
“It’s not necessary to mark everything up, but if you’re in your business starting out, it’s definitely a great way to add a cheat sheet. If you’re a site that is JavaScript-generated and you’re all client-side, pass along the cheat sheet and initial HTML to let Google know there is a payoff that this page will answer in intent once they go through the trouble of rendering it all.”
Is it a quicker way for a new brand to get Google more comfortable in what they’re about and what their business is?
“Absolutely, when you contribute to Wikidata or Wikimedia Commons, you’ll start adding in media elements like photos and videos to contribute to the understanding of an entity. Even, for example, with your home page - one of the boons is more traffic. When people want to understand the base construct of what a thing is, your site is the statement that says, ‘You’re not going to get this one, it’s well covered’.
Wookiees are part of the Star Wars universe, so every relationship between two entities is made up of a factual statement.”
Is it possible to automate this or does it need to be done on a manual basis?
“You can certainly try, but it’s a community. It’s similar to how Wikipedia was exploited with a pullback for people being able to edit it. It’s the same kind of trust with Wikidata. You don’t even necessarily have to start with your brand. Start with something that you are presenting, as a business, to be an expert on and contribute there.”
Is traditional SEO still something that’s very important for building entity trust?
“As we look at our use of heading tags, we’re semantically structuring the content of the page. That’s the same way that an entity is semantically structured and that product data markup is semantically structured. This is a repeating theme, both 1.30 and 3.0 are made of interlinking data structures.”
Does Google have any preferences in terms of content length, content type, etc., when it comes to giving more trust to what an entity is about?
“There are no limits or requirements on length. If it’s something super simplistic, they probably already know it and could tell you from search directly. If the factual data add that you provide can be summed up in a rich result, you’ll need a better page that’s more complete because no one will be flicking through yours.
If they had one question and it was answered just now, they won’t go to your site. Zero-click SERPs aren’t going away - they’re powered by entities, so you need to be at the trusted source that provides that content. We’re already seeing this change in how vaccine information was presented. MUM is powering those results and it’s helping to give contextual news results. It’s exciting and scary at the same time.”
How does an SEO articulate the financial value of doing an activity like helping to define the entity better, inserting schema on every single page, etc., instead of more traditional SEO activities to senior business leaders and business decision-makers?
“Because these are the ultimate links, you’ll have to submit any media to creative commons. Make an experiment and set KPIs. This is how you can define success, whether something will work out or not. You can share success on the first thing and then start building trust. However, SEO is a very esoteric field - a lot of the time it can just sound like word salad.”
Does that mean it’s better to just have a single domain to establish your reputation as an entity?
“When twelve clones are walking around, how do you know who the real one is unless there is a distinct value proposition, different audience, and a different tactic for engaging? Perhaps you’ve created top-level country code domains but also lost that domain property view that lets you see from one place.
Is everything working correctly? Are they all functioning the same? Are response codes returning the same thing? You need visibility to be effective and if your visibility is now spread out across all of these domains, how can you even spot when you’re cannibalizing your own content if hreflang links aren’t set up properly and all of these pieces are now fractured across the web?”
If you decide to change your brand and domain, is setting up permanent redirects from individual pages to the alternative individual page in the other domain enough to persuade Google that the entity has shifted? Are there other things that you need to do as well?
“There is some fun structured data markup you can put in there as well. If you have claimed your Knowledge Graph, you can update the information there. If you have a Google Business Profile location, that information should be updated there. One of the biggest markers will be part of the signals. They won’t just be hearing from the one site that your name changed. They’ll hear from five sites - from five trusted sources - that your business has changed its name.”
What shouldn’t SEOs be doing in 2023? What’s seductive in terms of time, but ultimately counterproductive?
“Adding more third-party tools to the head of the page. You need to protect the head. Even if the sales guy says it’s plug and play: ‘Just put the script directly above the opening tag and then magic elves will run across your web page and people will convert everywhere!’
They may not know they’re lying to you, so be kind, but also audit your page. Are you actually using everything that’s on there? Are you accurately collecting data? If you fire your analytics via Tag Manager through this mousetrap-style machine of mechanism and machinations, are you sure this was done right?
You need good visibility to make good decisions and you need good data to make good decisions. If you have a bobblehead site full of single points of failure and render blocking scripts, plus far too many A/B testing loads, you won’t need them on every page and you certainly won’t need twelve of them. Protect the head. The bobblehead is out, the svelte head is in.”
Jamie Indigo is a Senior Technical SEO at Lumar and you can find her at not-a-robot.com.