We Share, Inspire, and Celebrate Outrageously Successful Ethical Businesses and their Leaders

Google’s Core Topicality Systems Explained

August 6, 2024
roulette
For algorithms, it can be difficult to categorize vague and often unquantifiable notions like topics. Topics, trends, and genres have a way of blurring together when looked at too closely, but Google’s Search Off The Record podcast recently mentioned the existence of ‘core topicality systems.’ Here’s a brief rundown of everything we know about them.

Topics on the Internet

It’s not something we think about often, but topics play a vital, fundamental role in how we find online content. While topics can flow naturally in real life, the internet requires hard, quantifiable words and phrases to point you in the right direction. That’s what gives us keywords, which often use products and activities.

The names of industries and individual businesses are also topics and can intersect with more general labels like tech or retail. That’s to say nothing about branding and how that’s key to establishing a strong SEO profile. For example, take the iGaming industry where websites can belong to software, gaming, and gambling topics all at once. On those sites, services like Paddy Power’s roulette use live streams to broadcast interactive table games to their player base. In that example alone, the brand name, streaming, and roulette are also topics that could provide avenues for ranking.

Modern SEO experts know all there is to know about keywords, but topics are much broader and can be tricky to define sometimes. At any one time, a page on your site could belong to many different topics at once. So, how does Google detect them, and how do they decide which ones affect your ranking?

Core Topicality Systems Explained

That’s where Google’s core topicality systems come in. These systems were mentioned in episode 76 of the tech giant’s Search Off The Record podcast, titled ‘Measuring Search Quality with Data.’

The first mention of topics comes up in the first four minutes of the podcast. Google’s Director of Product Management Elizabeth Tucker is asked about user-reported satisfaction, specifically “granular ways” of measuring a satisfying search. In response, Tucker highlighted how topic management in search has improved over the years. She explained how user search quality varies – some use neat keywords while others tap their vague, conversational thoughts into the search bar. She says that keywords are easy to tie to a topic, unlike less specific searches.

She goes on to say that Google uses many different systems to rank queries where they belong. Perhaps the most publicized is their helpful content system which has been the tip of the spear for sweeping content policy changes throughout 2023 and 2024.

Core topicality systems

“Core topicality systems” come up at 11 minutes into the podcast. There Tucker goes into more detail about the query quality and how this relates to topic finding. She reiterates that using keywords makes topical categorization much easier, though adds that machine learning and the emergence of tech like natural language processing has helped a great deal. Tucker also highlighted how “the better we’re able to do this, the more interesting and difficult searches people do.”

The last mention of topicality occurs at 13 minutes, where Tucker refers again to Google’s transition away from an overbearing focus on keywords. She reveals that featured snippets were actually part of a larger move toward understanding topics in the context of longer, more complex search queries.

Unlike the helpful content system, there are no official branded core topicality systems outlined in Google’s resources. They aren’t widely publicized or explained further, and they are likely a bunch of smaller functions that Tucker described as one for the sake of convenience. As for how to make those systems happy, moderation seems to be the key. Don’t use a rigid, spammy, overly literal adherence to keywords but at the same time, don’t deliberately pose a challenge by making queries too long and vague.