What is BERT?
BERT went live the last week of October 2019 and marked the search engine’s biggest contextual algorithm update since RankBrain in 2015.
The acronym stands for Bidirectional Encoder Representations from Transformers and Google released a detailed report announcing BERT research back in May. If it sounds complicated, that’s because it is. BERT is Google’s latest machine-learning technique for natural language processing.
In other words, BERT helps Google’s bots understand how humans speak and what we truly mean when we say things or ask specific questions.
Google says that BERT impacts about 10% of all searches. Results for broad keywords like “office storage” won’t change much. However, long-tail searches with prepositions – like “cool storage for home office with small space” – should give the searcher much more relevant results.
Why BERT Matters More to Content Marketing Than You Think
1 out of every 10 searches may not seem like much in the grand scheme. However, if you rely on long-tail searches for traffic, BERT matters very much.
Not to mention, voice searches – which will make up 50% of all searches by 2020 – tend to include longer keywords rather than two-word phrases. People speak into their devices with complete sentences or questions – they don’t blurt out things like “car rental price Denver.”
BERT gives searchers much more accurate results. In terms of design and content marketing, this means Google bots are paying much closer attention to the context of each post and synonymous phrases.
Instead of checking off specific boxes by scanning your slug, headline, ALT text, and subheads, Google’s bots are looking at your content from a more holistic perspective.
In some cases, this could give smaller sites a major boost. Many websites with a high domain authority often outrank smaller sites for long-tail keywords even if the big site’s content isn’t super relevant. BERT could drastically flip that around.
3 Myths About BERT and Content Marketing
- You Should Optimize All Your Keywords for Stop Words
The BERT algorithm doesn’t specifically look for stop words like “to” or “from.” The algorithm has just improved how Google understands these words and interprets them to provide results.
- BERT Doesn’t Matter Too Much
Google processes over 78 thousand searches every second. 10% of over 3.5 billion searches each day impacts millions of search queries! That’s a massive impact.
- You Should Revamp Your SEO and Focus on Targeting Long-Tail Keywords
With BERT, Google wants to make its search results more relevant. Restructuring all your target keywords goes against everything BERT aims to solve.
If you have a solid domain authority and you’ve created content that answers searcher questions, you’re probably already ranking for plenty of long-tail keywords.
What BERT Means for Content Marketing and Creation
Marketers used to follow a plug-and-pray formula with keywords but that doesn’t hold up against human-like algorithms like BERT.
SEO and choosing keywords are still important – they tell the Google bots what your post is about. However, it’s much more important to focus on high-quality content marketing and creation.
Marketers should instead focus on a deeper level of keyword analysis. What type of questions are people asking? What problems are they trying to solve? How can I help them?
If you create interesting content and build your domain authority with backlinks, you’ll rank for plenty of natural language keywords without even really trying.
Conclusion
You can’t optimize your digital marketing specifically for BERT, but you can learn about your target audience and how to answer their questions. Human-like machine learning isn’t going anywhere. BERT is just the latest update in Google’s ever-growing AI program in an effort to improve search results.