Imagine if webpages were coded in a way that structured not just their form, but the meaning of their content. This would revolutionise the way that computer systems could use the information of the web. We could search for ‘taxonomy of penguins’, and Google could display a hierarchy of the penguin species, rather than just links to the pages containing these kinds of content.
It’s an exciting idea, because it would mean that Google can extract this information from webpages easily, and then devote its resources to displaying it in the best ways in response to your web searches.
But honestly, that ship has sailed.
The concept of the ‘semantic web’ was formulated in 2000 (Source), and yet it has never been implemented across the web. It’s ending up as one of those ideas that would have been so great if everyone had got on board. But really, what would have made everyone get on board?
How Google could have made the semantic web happen, but didn’t
Hundreds of thousands of web developers are not just going to start semantically coding their webpages without a clear motivation for doing so. In fact, ‘wait a second’ they thought, ‘why am I going to give out my website data to another service to display, when I want people to come to my actual website? This could undercut my website’s SEO!’
So who is going to motivate these lazy, selfish, rational web developers? Only the all-powerful Google, who could have provided SEO rewards that countered the negative SEO effects. Didn’t Google want to become the all-powerful, all-knowing semantic web deity?
Not if semantic code would look anything like this…
<section itemscope itemtype="https://schema.org">The <span itemprop="habitat"><span itemprop="species">little <span itemprop="animal">penguin</span></span> colony at <span itemprop="location"><span itemprop="city">Sydney</span>’s North Harbour</span></span> was <span itemprop="local-knowledge">a local secret for many years with residents reporting its existence from <span itemprop="date-range">around the 1940s and '50s</span></span>.This is psuedocode, but really it’s just an extreme version of the semantic code that is used currently. This kind of semantic markup was heralded as the next big thing on the web, although you can see how it quickly becomes extenuatingly difficult to write and edit when used on even a single sentence. Could a large part of the web be marked up like this? All practical reasoning says no.
So let’s just give up on the dream? No! It is in the making right now, except it doesn’t involve nearly as many markup tags.
How Google is making the semantic web happen, but it isn’t calling it that
Google is making the semantic web happen, and it’s going to be every bit as great as we thought it was going to be. We are going to be able to search in Google and get direct answers to our questions, in fact… we can do that right now for many queries. I.e. ‘how short is a little penguin’?
Google says:
30 – 33 cm.
Aww how cute.
But how is Google identifying all this data? Have websites finally started marking up their content semantically? Not overly, it’s because Google has started using AI.
AI is making the semantic web happen, but it’s not going to be called the semantic web; that’s a hackneyed 2000’s term. It’s going to be just be called AI, or some AI-related term that Google comes up with.
Google’s increasingly sophisticated AI is extracting more data, from less semantically marked-up code. It can identify the meaning in natural language text with an extraordinary degree of accuracy, and this helps with all aspects of indexing and ranking websites.
We are entering a future when Google and AI will be synonymous. Google will be a world-leading AI supergiant, and Google Search will be incredibly sophisticated at finding the exact answers that you need to your questions. The list of website links in Google Search will recede further away, and be replaced by more curated info by Google itself (drawing from the vast information of the web).
But there will still be a place for some semantic markup, and there always has been…
FACT FACT FACT
There are some key facts that Google’s algorithms, including its AI algorithms look for on your page. For a small business website, these include the business name, address, phone number, email address, etc. Hopefully it will be able to find these on your site … but how confident will it be in this finding? Even the smartest AI’s use fuzzy logic that uses probability to make decisions based on the degree of certainty. If an AI has to extract your phone number from a paragraph of text, it won’t be very certain. If it can find it in a neat table of key facts, it will be more certain. But it will be the most certain if you mark up this data semantically. The code won’t be overly complicated or long-winded, but it will improve your Google web presence.
Here is an example of small business details in a semantic markup that Google recognises called JSON-LD:
<script>
{
"@context": "https://schema.org",
"@type": "LocalBusiness",
"name": "Coffee Zone",
"address": {
"@type": "PostalAddress",
"streetAddress": "24 Morts Road"
"addressLocality": "Mortdale",
"addressRegion": "NSW",
},
"telephone": "9525 0414"
}
</script>This gives Google your business details in a language that Google’s robots can understand.
If you mark up your key business details in this semantic way, your business may appear in a card on the right side of Google searches, and this is better than being the top result on Google!
It’s a new reality where the top result on Google isn’t the most sought-after position anymore, and it’s a new future where the AI revolution is just beginning. For a website that includes similar semantic markup similar to this in the code, and also includes your social media profiles, star ratings, opening hours and more, you should have a look at our Jetsites. Find out more about how Jetsites are built to thrive now and into Google’s future →