The AI Revolution: BERT and Natural Language Processing
Google's BERT (Bidirectional Encoder Representations from Transformers) rollout in late 2019 and its continued refinement through 2020 marked what Google called "the biggest leap forward in the past five years, and one of the biggest leaps forward in the history of Search." According to Moz's BERT analysis, this represented the most significant change to search algorithms in recent years.
BERT affects approximately 10% of all searches, but its impact is far more significant than its percentage suggests. The algorithm fundamentally changed how Google interprets natural language, particularly in long-tail queries and conversational searches. Where previous algorithms processed search queries word by word, BERT considers the full context of a query--understanding how prepositions like "to" and "for" change the meaning of a search.
As noted by Search Engine Land's BERT coverage, this algorithm impacts one in ten searches but transforms how search engines understand complex queries.
How BERT Changed Search
- Bidirectional Processing: BERT processes queries from both left and right, considering full context
- Long-tail Keywords: Improved semantic understanding made long-tail targeting more important
- Question-Based Content: Natural question formats gained prominence in search results
- Intent Matching: Content must answer specific search intent, not just match keywords
The practical implication of BERT is straightforward: write content for users, not algorithms. As Google's Danny Sullivan stated, there's nothing specific to optimize for with BERT--the fundamentals of creating great content remain unchanged. This means focusing on natural language that reflects how people actually search, comprehensive topic coverage, and clear answers to your audience's questions.
For keyword research strategies that align with BERT's semantic understanding, focus on topic clusters and related concepts rather than isolated keywords. The rise of AI-powered search also underscores the importance of working with an AI automation agency that understands how machine learning impacts search visibility.
Core Web Vitals and Page Experience
In May 2020, Google announced that Core Web Vitals would become a ranking signal, officially integrating page experience into its ranking algorithm. According to Google's Core Web Vitals documentation, these metrics represent Google's effort to quantify user experience.
The Three Core Metrics
Largest Contentful Paint (LCP) - Measures loading performance. A good LCP score is 2.5 seconds or less, indicating that the main content loads quickly and users can begin engaging with your page without frustrating delays.
First Input Delay (FID) - Measures interactivity. A good FID score is 100 milliseconds or less, indicating that pages respond quickly when users try to interact with buttons, links, or other interactive elements.
Cumulative Layout Shift (CLS) - Measures visual stability. A good CLS score is 0.1 or less, indicating that page elements don't shift unexpectedly during loading, which creates a smoother and more predictable user experience.
Technical Implementation
For LCP (Loading Performance):
- Optimize and compress images using modern formats like WebP
- Implement lazy loading for below-fold content
- Minimize render-blocking CSS and JavaScript
- Use content delivery networks (CDNs)
For FID (Interactivity):
- Break up long JavaScript tasks into smaller chunks
- Defer non-critical JavaScript execution
- Minimize main thread work during page load
For CLS (Visual Stability):
- Always include width and height attributes on image elements
- Reserve space for advertisements and dynamic content
- Avoid inserting new content above existing content
Core Web Vitals are part of a broader technical SEO strategy that ensures search engines can efficiently crawl and index your site while delivering excellent user experiences. If your site has SEO issues related to performance, addressing Core Web Vitals should be a priority.
Strategies for demonstrating expertise, authoritativeness, and trustworthiness to Google
Expertise Signals
Author bylines with credentials, author biography pages, citations to authoritative sources, and original research demonstrating genuine knowledge.
Authoritativeness Signals
Backlinks from reputable industry websites, mentions in publications, professional certifications, and positive reviews from clients.
Trustworthiness Signals
HTTPS security, clear contact information, privacy policy, transparent advertising disclosures, and consistent information across the web.
Voice Search and Structured Data
With over 120 million smart speakers in the US and 35.6 million voice searches occurring monthly, voice search has become a significant channel for SEO. According to Moz's voice search research, this channel requires specialized optimization strategies.
Voice Search Characteristics
- Conversational Queries: Longer, more natural language patterns
- Question Format: Often takes the form of complete questions
- Local Intent: Strong geographic component to many queries
- Immediate Answers: Demand for concise, direct responses
Schema Markup for Featured Snippets
Schema markup helps search engines understand content and improves featured snippet chances:
- Organization Schema: Brand information and entity recognition
- Article Schema: Blog posts and news content
- Product Schema: E-commerce product pages
- FAQ Schema: Question and answer content
- HowTo Schema: Instructional and tutorial content
- Speakable Schema: Voice-optimized content for Google Assistant
Implementing structured data through schema markup is an essential part of modern SEO that helps search engines parse and display your content effectively across different search features. For content teams looking to understand the broader landscape, exploring AI SEO statistics can provide additional context on how artificial intelligence continues to shape search behavior.
Voice Search by the Numbers
120+
Million smart speakers in the US
35.6M
Voice searches per month
10%
Searches affected by BERT
2.5s
Target LCP for Core Web Vitals
Entity-Based SEO and the Knowledge Graph
Google's shift toward entity-based understanding represents a fundamental change in search. Where traditional SEO focused on keywords (strings), modern SEO must understand how Google identifies and evaluates entities. According to Moz's entity-based SEO guide, this shift represents the evolution from optimizing for strings to optimizing for things.
What Are Entities?
An entity is "a thing or concept that is singular, unique, well-defined, and distinguishable." This includes:
- People: Authors, experts, public figures
- Organizations: Companies, nonprofits, government bodies
- Places: Business locations, event venues
- Concepts: Topics, ideas, methodologies
- Products: Physical or digital offerings
Building Entity Authority
- Claim and verify Google Business Profile
- Create Wikipedia or Wikidata entries for notable entities
- Build consistent NAP citations across the web
- Implement schema markup for entity relationships
- Earn mentions from other established entities
Entity-based SEO works hand-in-hand with content marketing services to establish your brand as a recognized authority within your industry and topic areas. Understanding how off-page SEO factors into entity authority can amplify your efforts in building recognition across the web.
Semantic Search and Topic Authority
Google's ability to understand topic relationships has grown dramatically, reducing the importance of exact-match keywords while increasing the value of comprehensive topic coverage. According to Moz's topic authority guide, semantic connectivity has become a key ranking factor.
From Keywords to Topics
Traditional keyword targeting has evolved into topic authority building. Google's algorithms now prefer:
- Comprehensive Coverage: Thorough exploration of a subject area
- Topic Clusters: Interconnected content around core themes
- Pillar Pages: Definitive resources on broad topics
- Internal Linking: Clear relationships between related content
Content Cluster Architecture
Pillar Content: Comprehensive, long-form resources covering broad topics (10,000+ words)
Cluster Content: Supporting articles covering specific subtopics in depth (2,000-4,000 words)
Internal Links: Clear connections between pillar and cluster content
Topic Depth: Evidence of genuine expertise through thorough coverage
This semantic approach to SEO aligns with our content strategy services that build topic authority through interconnected content ecosystems. Proper keyword analysis forms the foundation for identifying the right topics to target in your cluster strategy.
Frequently Asked Questions About 2020 SEO Trends
What was the most significant SEO change in 2020?
The integration of Core Web Vitals into Google's ranking algorithm represented one of the most significant changes, making technical performance and user experience official ranking factors alongside traditional signals like content quality and backlinks.
Can I optimize for BERT?
No. BERT is a natural language processing algorithm that rewards high-quality, user-focused content. Google's own guidance emphasizes writing for users rather than optimizing for algorithms. The best approach is creating comprehensive, well-written content that genuinely answers searcher questions.
How do Core Web Vitals affect rankings?
Core Web Vitals (LCP, FID, CLS) are part of Google's page experience signals. While not the only ranking factor, pages with good Core Web Vitals scores may have an advantage in competitive search results, particularly on mobile devices.
Is E-A-T a direct ranking factor?
E-A-T itself is not a direct ranking factor, but it is used by Google's quality raters to evaluate content quality. Demonstrating expertise, authoritativeness, and trustworthiness remains critical for SEO success, especially for YMYL topics.
How important is schema markup for SEO?
Schema markup helps search engines understand content and can enable rich results in SERPs. While not a direct ranking factor, it improves how content is displayed and can increase click-through rates through rich snippets, knowledge panels, and featured answers.