These thoughts I mention all centre around topical targeting, machine learning, snippets and the tendencies of the SEO industry to take things quite literally. (Shocking video!) 😜
Initialising brain dump…
At MozCon, Dr Pete spoke about the concept of keyword research and content production based on topics and context. As Google machine learning understands topics – the relationship between keywords and phrases and inherent intent – we as SEOs have to free ourselves from the shackles of keyword specific targeting and think of the bigger picture. It’s like something out of a dream: produce content that means something to real people, rather than content for search engines. Long gone are the days of one page per keyword.
Dr Pete’s awesome talk was hammered home even further with the recent Whiteboard Friday about Rank Brain. Rank Brain, Google’s algorithmic toy, weights ranking signals based on topic and intent. Now, some queries are ranked more heavily based on links and some more on freshness, for example. On the face of it, it seems unrelated but it all hangs together by a faintly visible thread. Look closely, and you’ll see this thread is everywhere, in every small or big Google announcement, algorithm update (like Penguin 4.0) or interpretation of search engine weather forecasts.
Most important detail of the announcement is the change to how it operates (ML-based consideration of ranking signals, more like RankBrain) pic.twitter.com/eA1Lo37A14
— Rand Fishkin (@randfish) September 23, 2016
The thread is the presence of machine learning. The beautiful thing is that machine learning allows Google to be more human. Search results are so truly focused on the searcher’s intent and put the results into the wider context of a ‘concept’. This enormous shift in Google’s capabilities forces us into an SEO world that truly focuses on content concepts, rather that optimising for very specific keywords and phrases. Because that’s not how the human mind works – we don’t think in keywords.
And here is where it all falls down. Another talk as MozCon focused around the upcoming gold rush for featured snippets. Rob Bucci gave very insightful and actionable advice on how to be a contender for the ever more prevalent featured snippet. The problem with the way featured snippets work, currently, is that they still do work with quite literal phrases, often questions. Like creating a step by step list for “How to pan for gold” and separate content for “How to do gold panning step by step”. So in order to get one of these highly sought after #0 slots SEOs are yet again playing the keyword matching game. We have been advised to research and target very specific phrases, produce content in a very specific way to match those phrases, all in order to fit into a space/format that Google has provided for us.
For me this feels way too much like opening another door to spammy, short sighted SEO tactics, like creating a myriad of pages or sections that are highly optimised for a specific query, just when we thought that door was forever closed.
But the current/imminent gold rush for snippets might just be the last “easy” opportunity for SEOs to game Google. It can’t be long until machine learning capabilities slam that door shut, too. Now they understand concepts around keywords, relationship and intent they surely will soon be able to do the same for longer phrases and questions. Which means that those featured snippets could be taken from a site that gets the content and intent exactly right and simply dynamically framed with a fitting title to match the searcher’s question.
So my quibble is whether SEOs should create content right now to specifically target those questions and phrases to secure their very own #0 ranking? Or should we play the long game and expand our research further to provide the best result from a contextual perspective?
Has anyone tested a content grouping snippet strategy yet? I’d say if it doesn’t work yet, it won’t be long until it does. That’s where I am putting my gold. 💰