Google's AI vs. lazy content: Guess who wins

Be the first to comment

The search giant's AI is too smart for bad content ... and it's getting smarter. Here's how to look good in front of the machines, says Critical Mass strategy director

The days of Artificial Intelligence (AI) are upon us. We have entered an age of machines that can learn from experience, of algorithms that propagate smarter algorithms as they collect more data. While a Terminator-style awakening of Skynet is not yet at hand, advancements in machine learning are introducing possibilities formerly known only in the realm of science fiction. 

These breakthroughs are poised to dramatically improve user experience across an array of channels. But digital marketing mavens of all stripes will likewise have to contend with revolutionary changes in SEO and content strategy. They’ll need to understand what these changes driven by machine learning entail — and what it will take to adapt, to succeed and to thrive in a rapidly evolving, uncertain future. 

So, do you know what you’re up against? 

Google’s pursuit of search perfection, that’s what. It was Google, after all, that cracked the whip in 2011 and unleashed the "Panda" algorithm, beating back a range of Internet plagues — content farms, sites with high ad-to-content ratios and thin content in general. Yes, the algorithmic change hit a few marketers and content publishers like a meteor, but users loved it, and marketers learned how to adapt to the new ranking signals.

But now, meet RankBrain, purported to be Google’s third most important rank signal (out of hundreds). Greg Corrado, a senior research analyst with Google, explained how it works to Bloomberg:

RankBrain uses artificial intelligence to embed vast amounts of written language into mathematical entities — called vectors — that a computer can understand. If RankBrain sees a word or phrase it isn’t familiar with, the machine can make a guess as to what words or phrases might have a similar meaning and filter the result accordingly, making it more effective at handling never-before-seen search queries.

RankBrain represents one more giant leap beyond the mere recognition of search inputs; it proactively deduces what search entries actually mean. (Deducing meaning is supposed to be one of those things that make humans human). Oh, and RankBrain was 14% better at guessing page rank results than the Google engineers who made it. We’ll let that sink in for moment. In four years, Google algorithms progressed from "very helpful" to "slightly superhuman."

The winter of our dissed content 
And oh how expectations will rise. Machine learning will allow Internet users to search ever more semantically, and as a result their demand for miraculously relevant content will increase dramatically—a human need, in a human moment, fueled by a mix or curiosity and impatience. 

Meanwhile, search engines will classify more and more content as "run-of-the-mill" and fence it off. So if figuring out how to make your content connect with specific user needs in deeply contextual moments represents one side of the SEO Holy Grail, the other side of that Grail is top-notch content — or what is now being referred to as 10X content.

First coined by search guru Rand Fishkin, 10x Content refers to content that is 10 times better than the best result that can currently be found in the search results for a given keyword phrase or topic. A beautifully simple concept. A mindset and a roadmap at once. 

So how do you create it? Start with the right questions. Ask yourself, what micro moments do you need to win at? Compile a list and conduct a search along each of those questions. See what you come up with in the results.

Then ask yourself:

  • What sorts of user experiences are provided? 
  • How mobile friendly is the content?
  • How fast does it load? 
  • What is the quality of the design we are up against?
  • What is the quality of the copy?
  • How much work is required of the user to be able to get the information?
  • How thorough is the content? Does it meet all the needs a user might require? What information is missing?

Next, score your content against the same criteria. From this, it is possible to determine a delta between you and your competition. Once that delta has been determined, plan to create content that is 10 times better than the best results in the current ranking. 

The rise of content marketing over the last few years has made the competition fierce. But that is the new bar to which we will be measured and ranked. And if winning at the above questions provides a key competitive advantage, it will indeed be a huge advantage in search, in social and on the Web as a whole. The strategic approach doesn’t need to be ove-complicated.

Invest in the change
As machines do more thinking, we need to optimize the way we spend our marketing dollars. 10x content can be difficult and even expensive to create — but the investment should prove worthwhile. Even if we buy our way to the top of the search results, we’re doing something akin to putting our content on life support, rather than helping it thrive in a changing digital ecosystem. 

At stake here is an existential showdown between brand content as something "advertised" and something "experienced." As the rise of Digital Experience Design suggests, the move from "advertising" paradigms to "experience" paradigms is already under way, and this movement is proving prescient. By connecting your users (or your digital customers) with a better content experience, you have already gone a great ways toward producing the very thing that machine-learning search algorithms will increasingly favor. That’s where your time, money and business plan belong.

So as the smart machines rise, rise with them: Up your content game. Or get terminated. 

Mark Silverman is Critical Mass strategy director.


The latest work, news, advice, comment and analysis, sent to you every day

register free

Campaign Jobs

Thousands of jobs across advertising, creative, marketing and media

Trending on Campaign


The Hub 
Marketing Tech News