The problem modern SEOs face today is one of falling organic traffic, as the amount of organic traffic that Google sends to other sites has declined nearly 15% between 2016 and 2019.

Conductor partner Talkwalker, in collaboration with Trust Insights, published an original research paper examining this conundrum, and solving it with the power of artificial intelligence and social media conversational data also know as social media mentions. 

In brief, Google’s algorithms are trying to deliver the best results based on human conversation, at scale. Instead of using SEO techniques of the past like keyword stuffing, or machine-generated answers, the best SEO practice is–what, to some extent, it’s always been–original content that speaks to humans.

But speaking to humans is hard. More specifically, finding out just what humanity is speaking about at a given moment, and where that conversation is likely to turn, is hard. Curating these conversations is really only a possibility with a neural network of, let’s say, smart speakers. 

That means it’s inaccessible to all but a handful of companies. It also puts mimicry of Google’s AI modelling, a popular technique of the past, out of reach. 

Marketers and SEOs should fear not, however. 

A more out-of-the-box solution is to source these conversations from social media, a playground and cacophony of voices and self-selecting communities diving deep into the details of their passion and sharing tips, wisdom, and promulgating any given topic’s conversation. Doing so will allow you to capture, measure, and organize these mentions in ways that will help you establish insights and findings from the vast wealth of social media data.

Every day on social media billions of conversations on millions of topics are taking place. 

As these conversations occur, they grow in network strength – more people get pulled in, and discuss with even more people. 

As curiosity or passions are ignited on social media, search volume on these subjects follows.

The paper finds that there is a 9-10 week lag, but correlation, between social media mentions and the ensuing search results for the topic searched. This may differ depending on industry, topic, and of course broader cultural trends. 

As a rough guide you can build this type of time lag into your content calendar. 

Using a basic keyword in which you have some knowledge (your domain topic, essentially) you can mine the social media data for conversations around the topic. 

Bi-grams display the frequency with which two words appear together. While search volume statistics can tell us just how frequently a term like “accounting software” is searched for, they give us little direction for what else is searched. 

What would a third, or fourth word most likely be after those two words? 

Using the bi-grams and tri-grams feature of Talkwalker’s Quick Search tool gives us a real-time look at the phrases being used in actual human conversation on social media networks right now. 

Accounting software is a massive industry, filled with nuance. Google wants search results to reflect this breadth of conversation in order to deliver the best human experience possible. 

The simplest way to make sure that your content is both of those things, rich and nuanced, useful and satisfying, is simply to create content that reflects this.

Caption: using bi-grams in Quick Search we can see a robust list of terms related to “accounting software”.

Filtering to create a topically relevant list, we’re left with a much richer and more varied list of two-word phrases that searchers of “accounting software” are also discussing on social media. These secondary terms can often be much easier to rank for, and can see a ton of valuable traffic in their own right.

The paper finds a difficulty-discounted achievable search volume of more than 60% for this list of related keywords.

This is just one keyword, in one business niche. But still, that’s a substantial improvement that should capture every SEO’s attention. 

The paper goes into a deeper explanation of n-grams and how to use AI to exploit them. 

It also includes powerful insights into visual SEO strategies. Emulating the style and format of the most successful conversants in any given topic’s social media space is an easy way to rank up on image SERPs. 


The above strategy really only hints at the possibilities that are achievable with this approach. 

Download Powering modern SEO with social media data to dig deep into how conversational clustering can uncover rich topics your brand can occupy, as well as niche communities who will find your content invaluable.

If your SEO senses are tingling, here are the keys to remember for human-powered SEO success:

  • Google wants to deliver the best search experience…for humans. It uses AI to iterate this.
  • Our methods of the past like keyword-stuffing or machine-optimized content are not the best experience for humans. Google recognizes –and penalizes– this.
  • A powerful and adjacent re-creation of the human search experience is happening on social media, often before it hits the search bar. 
  • Your can derive insights from this data to optimize your brand for the human experience using these adjacent terms, which reflect what people are really talking about. 
  • Achievable search volumes in this adjacent sphere are simply too eye-popping to ignore.
  • Create content for humans! That humans will want to read, will enjoy, and will find useful enough to share. 

Comments are closed.