April 28, 2020
Google sent shockwaves through the digital world in October, 2019 when they announced BERT’s arrival. What is BERT? It stands for Bidirectional Encoder Representations from Transformers. What does that mean? It’s the most comprehensive update to Google’s search engine since Google was created, and is powered by machine learning. The update is to help Google to understand the context of a query rather than in a string from one word to the next. Rather than simply matching words, BERT can understand what users are asking to provide more relevant results.
BERT stayed in an open-source testing stage for a year before its official introduction as Google determined if its presence would do what they wanted it to: refine search query results.
Instead, BERT shattered the algorithm.
Affecting at least one in 10 search queries, the overhaul strips away the tricks and strategies digital and search engine strategists have relied on since the inception of search engines themselves. Keyword stuffing was already a piece of history before 2019 but genuine content won by a landslide in the waning months of the decade.
While nothing in digital marketing is set in stone, Google’s embrace of artificial intelligence signals a seismic wave coming for the industry. The 2020s will be witness to the figurative rise of machines.
You might have noticed the relaxed nature of this article. In the past, an article reporting on Google algorithm changes might have employed a tactic that worked something like this: It starts with an introduction, then a copy-paste explanation of what Google search is, followed by a brief history of Google as a company. Finally, you’d find a paragraph or two of the actual news, topped off with a reiteration of the previous paragraph for good measure.
This was a shoddy practice meant to game an increasingly sophisticated algorithm. Google had already discovered a solution to eliminate keyword stuffing, so it was only a matter of time until a company of their sophistication would find a way to destroy the algorithm and replace it with something even more customizable.
Google’s public search liaison Danny Sullivan put it boldly when asked how sites can adapt their search engine optimization strategy to the new system: “There’s nothing to optimize for with BERT, nor anything for anyone to be rethinking. The fundamentals of us seeking to reward great content remain unchanged.”
BERT results affected one in 10 searches in October. Under the basic principles of machine learning, this has or will soon grow to affect even more results, and who knows how far it could go?
Many have seen their site traffic increase while others are reporting steep drop-offs. No matter what happens in 2020, strategists must readjust away from chasing the algorithm and toward learning to live with the machines.
The easy conclusion to make is that search engine ranking is now outside the reach of strategy, search engine optimization is as dead as keyword stuffing. View this as a partial truth. Optimization is an endangered practice, but strategy is more important than ever.
This will place the onus on quality content to carry the journey from casual visitors to conversions. No matter the size of the brand, establishment, or institution, the different facets that inform digital strategy—content, analysis, marketing, and much more—must work harmoniously in order to ensure a smooth transition as technologies like BERT continue to advance in the coming years.
As with any industry’s leader, Google’s choices will soon trickle down into most if not all facets of digital marketing, if they haven’t already. Look no further than Google’s closest competitor, Microsoft’s Bing. Bing began experimenting with new artificial intelligence implementations in 2018, around the time that BERT entered into open source testing. YouTube’s confidence in their complex system using title, description, and contents of the video to learn has been so successful the platform has recently deemphasized the need for users to add keywords when uploading new videos.
Looking to the future, risk and reward of ad bids will become more granular and audience targeting will become even more hyper-focused. What we wish to express in articles or ad copy won’t be held back by algorithms.
Artificial intelligence is forcing us to remember what it was like to appeal to humans.
Of course, all of this can seem overwhelming. Digital strategy changes daily—sometimes hourly—and requires a full staff worth of knowledge and ideation. In the absence of that, let’s start a conversation on how Stamats can help grow your digital impact.