Supercharged Rankbrain Content (the rise of medium tail)

Rankbrain does what Hummingbird was supposed to do. Hummingbird itself replaced Caffeine. And Caffeine was built on the success of two other … OK I’ll stop. The point is that Google has a long history of rolling out algorithm updates that ‘promote or punish’ when it comes to content. But hasn’t the Google mantra always been to stay safe by producing ‘original unique content’? Well, yes. However, is it also safe to say that there are ways to make sure that your content is viewed as favourably as possible in the eyes of Google? Absolutely.

In this article, we will be discussing how the introduction of Rankbrain has led to a rethink for content creators. Spoiler alert: longtail is out, medium tail is king. Let’s start at the beginning.

Shifting Goalposts – Ranking Through the Ages

About ten years ago, I shared a cab ride with some friends. The driver asked how we’d found the taxi company’s phone number, and someone answered that we’d searched online. “That’s why we’re called AAA taxis”, he replied, “puts us top of Google”. Cringe. He’s probably still out there now, advising companies to rebrand with ‘Aardvark’ at the start of their name. In reality, of course, things are not only very different, but the playbook is occasionally re-written.

Here’s a brief history of content related algorithms:

Caffeine (2010)
Google Caffeine was loosely linked to the 2004’s ‘Brandy’ and 2005’s ‘Big Daddy’ algorithms. Collectively, these updates aimed to increase indexing speed/accuracy through Latent Semantic Indexing (LSI) – i.e. understanding the meaning behind search terms as a whole instead of paying attention to keywords on an individual basis.

Hummingbird (2013)
Hummingbird changed everything. For the first time, Google gave a nod towards the fact that the quality of content under Caffeine was still suffering from keyword-stuffing as content creators battled for ranking. Hummingbird introduced long tail search, which improved LSI (known as ‘Entity Search’). This meant that one well written piece of content would theoretically rank for many related search terms, even where the words used in search did not appear in that order on the page.

Sounds great, right? Yes. It does. But there was a problem looming. Content creators took Hummingbird and ran with it. Instead of trusting Google to fulfill the promise of improved LSI, the introduction of longtail search instead sent the creation of near-duplicate pages into overdrive. An abundance of landing pages began to appear, each attempting to rank for a longtail search term and each linking back to a parent web page. This ‘flooding’ tactic aimed to dominate the first page of search for all related longtail search terms by creating a funnel of landing pages. Ultimately, the usefulness of search results was being gradually devalued, where each potentially useful website was effectively closed off from searchers by its own wall of landing pages.

Rankbrain (2015)
All previous content related algorithms followed a strict tick-box style procedure when assessing page rank. This procedure was hand coded by Google engineers – and they began to realise that they simply couldn’t keep up. Instead, Rankbrain introduced ‘machine learning’, which updates in real time based on three things: content semantics, search query semantics, and the obvious usefulness of any given result based on user interaction (i.e. whether visitors stuck around to read the content).

So what does this mean for content creators? Carry on as usual with content tailored for longtail search queries and wait for the rewards? Not exactly…

The Growing Irrelevance of Longtail

Let’s begin with a quick definition: longtail refers to ‘question’ style search queries, where a query is typed out in full (e.g. “How many dwarves lived with Snow White” as opposed to “dwarves snow white”).

Longtail typically makes use of top level search terms in the hope that the content will also rank for related lower volume search terms. Going back to the Snow White example, we could choose to title a page “Snow White – How Many Dwarves Can You Name?” and rely upon Rankbrain to pull in all related searches with lower search volumes. However, using the top search terms leaves no room for Rankbrain to pull in search queries with GREATER search volumes as well as lower search volumes. This is where subtle use of medium tail keywords can make all the difference in ranking on page 1.

How to Use Medium Tail Keywords

Begin by selecting relevant medium value keywords. Instead of tailoring an entire page of content around a longtail high value search term, allow the content to breathe with inclusive information that is useful to the reader. For example, going back to Snow White and the Seven Dwarves, include a section on the Queen, the Witch, the Prince, and perhaps Walt Disney himself. These related keywords garner significantly lower search volumes per month than ‘Snow White’, but the usefulness of such inclusive content should trigger greater user interaction and increase dwell time.

Medium tail therefore removes the reliance on users to search for very specific terms.
In practice, this should result in the article gaining a cumulative ‘gravity’ for a variety of related search terms, which Rankbrain will reward with a higher ranking.

👇 Like what you read? Share what we said! 👇

About Ben Atherton

Ben Atherton

Ben is a grammar guru with a passion for the written word. Upon graduating in 2008, he trekked across Europe and produced his first travel book. A career in SEO copywriting for the law, music, and food industries followed, before recently spending a year in the role of magazine editor for a national publication.