AI Search in 2026: What 2020's SEO Predictions Got Right

Voice Search Predictions That Never Materialised

The widely-cited prediction that "50% of searches would be voice by 2020" was based on a misquoted 2014 Baidu statistic that specifically referred to voice or image searches in the Chinese market only. Current data reveals voice search has plateaued at approximately 20.5% of global internet users, with peak usage occurring in Q2 2022 at 22.5% before declining.

Smart speaker ownership reached 35% of Americans aged 12+ (NPR/Edison Research 2022)—far below the 75% projected for 2025. Amazon's Alexa division reported a $10 billion loss in 2022, demonstrating that voice commerce hasn't delivered profitable outcomes. The practical implication for SEO strategy: focus on featured snippets and local optimisation rather than dedicated voice search tactics.

What the 2020 article got right: Google's investments in natural language processing would fundamentally change search behaviour. The mechanism, however, wasn't voice—it was AI-generated answers that synthesise information directly in the search results.

Zero-Click Searches Have Increased Dramatically

The 2020 article cited Jumpshot data showing 49% of searches resulted in no click. Within weeks of that article's publication, Jumpshot shut down following a privacy scandal. SparkToro's July 2024 study using Datos (which acquired Semrush's clickstream capabilities) found the situation has worsened considerably: for every 1,000 US Google searches, only 360 clicks now go to the open web.

The data reveals stark differences by device. Mobile zero-click rates have reached 77%, whilst desktop remains at 46.5%. EU rates (59.7%) slightly exceed US rates (58.5%). AI Overviews accelerate this trend further—Seer Interactive's September 2025 study found organic click-through rates dropped 61% for queries showing AI Overviews.

For B2B marketers targeting personas like Marketing Directors at SaaS companies, this demands a strategic shift. Answer Engine Optimisation (AEO) focuses on getting your brand mentioned within AI-generated answers, not just ranking to earn clicks. Whitehat's approach combines traditional SEO with AEO to ensure visibility across both paradigms.

Google's Algorithm Evolution from BERT to AI Overviews

The 2020 description of BERT as "RankBrain 10.0" accurately captured its significance as Google's shift toward AI-powered language understanding. What followed exceeded expectations. MUM (Multitask Unified Model), announced May 2021, was described by Google VP Pandu Nayak as "1,000x more powerful than BERT"—trained across 75 languages simultaneously and understanding text, images, video, and audio.

The March 2024 Core Update represented Google's "biggest update ever," integrating the Helpful Content System directly into core ranking. This 45-day rollout specifically targeted scaled content abuse and "parasite SEO." Notably, the September 2023 update removed the requirement that content must be "written by people"—shifting focus to quality regardless of AI involvement.

AI Overviews launched May 2024, powered by a custom Gemini model. They now appear in 30-60% of US queries depending on measurement methodology, primarily for informational searches (88-91% of appearances). Position #1 organic click-through rates dropped from 28% to 19%—a 32% decline. This validates why businesses need integrated HubSpot and SEO strategies that capture attention across multiple touchpoints.

E-E-A-T Has Become More Important Than 2020 Suggested

The 2020 claim that E-A-T was "not a ranking factor" and "not a thing to worry about" was technically accurate but practically misleading. Google officially added "Experience" in December 2022—just 15 days after ChatGPT's launch—creating E-E-A-T. This timing wasn't coincidental: Google needed to differentiate human experiential content from AI-generated content.

Google's Danny Sullivan confirmed in February 2024: "It's not a ranking factor. It's not a thing that's going to factor into other factors." However, the same documentation states: "Using a mix of factors that can identify content with good E-E-A-T is useful. Our systems give even more weight to content that aligns with strong E-E-A-T for topics that could significantly impact health, financial stability, or safety."

The March 2024 Core Update made E-E-A-T more influential by embedding the Helpful Content System into core ranking. Industry consensus is now clear—as Ahrefs states: "A lack of E-E-A-T can hurt not just one page, but your entire site's visibility." The framework now determines visibility in both traditional search and AI Overviews.

AI Search Engines Have Changed Everything

This represents an entirely new category that 2020's predictions couldn't have anticipated. ChatGPT dominates with 800 million weekly active users as of late 2025, having doubled from 400 million in February 2025. The platform processes 2.5 billion daily queries (29,000 per second) and ranks as the 5th-6th most visited website globally. OpenAI's revenue reached $10 billion ARR by June 2025.

Perplexity AI has grown explosively, reaching 780 million monthly queries by May 2025—up from 230 million in August 2024 (239% growth). Gartner's February 2024 prediction stated: "By 2026, traditional search engine volume will drop 25%, with search marketing losing market share to AI chatbots." Current data shows 37% of US adults now start searches with AI tools instead of traditional search engines (Eight Oh Two, 2025).

For UK businesses, this shift is accelerating. IAB UK data from October 2025 shows 22.5 million UK people now use AI tools, with ChatGPT commanding 84.11% of UK AI traffic. Whitehat's AI consultancy services help B2B companies navigate this transition, ensuring visibility across traditional and AI search channels.

Topic Clusters Remain Valid But Require AI Optimisation

The 2020 article advocated for topic clusters and pillar pages—a methodology that remains recommended in 2026 but has evolved significantly. Google officially introduced "topic authority" in May 2023, and Search Engine Land notes: "Topic authority is critical. If you're good at establishing topic authority, you will show up prominently in the SERP."

HubSpot continues supporting the topic cluster model, with their original experiments showing "at least 50% more organic search results from the moment they started clustering content by topic." The approach now requires "dual optimisation"—content must be optimised for both traditional rankings AND AI citation.

Key adaptations for 2026 include: leading with 40-70 word concise summaries that AI can extract; using descriptive headers (research shows question-style headers underperform with 3.4 vs 4.3 citations); implementing FAQ schema; and demonstrating E-E-A-T signals through author credentials and expert quotes. This approach, which Whitehat calls Answer Engine Optimisation, ensures content serves both human readers and AI systems.

Link Building Has Been Officially Downgraded

The 2020 claim that links are "the foundation of SEO" requires significant qualification. Google officially changed its documentation in March 2024, downgrading links from an "important signal" to just "a signal when determining the relevancy of pages." Gary Illyes stated explicitly: "Links are important but people overestimate their importance... not in the top three ranking factors."

Empirical data still shows correlation—Backlinko's study of 11.8 million search results found the #1 result has 3.8x more backlinks than positions #2-10. However, Ahrefs' study of 1 million keywords found correlation between links and rankings has decreased since 2019, dropping to 0.22-0.24. The interpretation: Google's AI systems (BERT, MUM) can now assess content quality directly, reducing dependence on external link signals.

The practical shift: quality over quantity matters more than ever. A single high-quality editorial link outweighs hundreds of low-quality links. Digital PR, original research, and relationship-based linking have become the recommended approaches. Whitehat's SEO programmes now prioritise authority-building through content that earns natural citations rather than traditional link-building tactics.

SEO Tools Have Transformed Since 2020

The 2020 article recommended tools that have undergone significant changes. Jumpshot shut down January 30, 2020—within days of the article's publication—following a privacy scandal. AnswerThePublic was acquired by NP Digital (Neil Patel's agency) in May 2022 and now integrates with Ubersuggest. AlsoAsked.com remains operational, offering PAA data across 190+ regions.

Major new AI-powered tools have emerged: SE Ranking's AI Overviews Tracker monitors AI visibility for keywords; Frase.io launched GEO features for dual optimisation; Surfer AI Tracker monitors brand mentions across ChatGPT, Gemini, and Perplexity; and HubSpot released its AI Search Grader—a free tool analysing brand performance in AI search.

The established platforms have evolved significantly. Semrush added ChatGPT Search position tracking and visibility tracking across AI platforms. Ahrefs launched AI Content Helper but faced criticism for significant price increases. The tool landscape now requires capabilities for tracking visibility in both traditional search and AI-generated responses.

What B2B Marketers Should Do Now

The 2020 article's core insight—that Google's advancements in natural language processing would transform search—proved accurate. The mechanism differed from predictions, but the conclusion stands: SEO success requires understanding how search engines interpret intent, not just optimising for keywords. In 2026, that understanding must extend to AI systems beyond Google.

For Marketing Directors at B2B SaaS companies dealing with HubSpot attribution challenges and pipeline pressure, the strategic implications are clear: invest in content quality over quantity; structure content for AI extraction with clear, direct answers; build E-E-A-T signals through demonstrable expertise; and monitor visibility across both traditional SERPs and AI platforms.

The businesses that will thrive are those treating this transition not as a threat but as an opportunity to differentiate through genuine expertise and helpful content. Whitehat's integrated SEO and HubSpot approach helps B2B companies build sustainable visibility across all channels—traditional search, AI search, and direct demand generation.

Frequently Asked Questions

Is voice search still important for SEO in 2026?

Voice search has plateaued at approximately 20% of users rather than becoming dominant as predicted. Focus instead on featured snippets and AI Overviews optimisation—40.7% of voice answers come from featured snippets anyway. Optimising for clear, extractable answers serves both voice and AI search simultaneously.

How do I optimise content for ChatGPT and AI search?

Structure content with direct answers in the first 40-60 words, use descriptive (not question-based) headers, include statistics with source attribution, and ensure AI crawlers (GPTBot, Claude-Web, PerplexityBot) can access your content via robots.txt. Brand mentions should appear in extractable passages that AI can quote.

Are backlinks still important for ranking in 2026?

Yes, but less than before. Google officially downgraded links to "a signal" (not "important signal") in March 2024. Quality now matters far more than quantity—a single editorial link from an authoritative site outweighs hundreds of low-quality links. Focus on earning links through genuine expertise and original research.

What percentage of searches result in zero clicks?

Currently 58.5% in the US and 59.7% in the EU, up from 49% in 2020. Mobile zero-click rates are even higher at 77%. AI Overviews accelerate this—organic CTR drops 61% when AI Overviews appear. This makes brand visibility within AI responses increasingly valuable.

Should I still use topic clusters and pillar pages?

Yes—topic clusters remain effective for establishing topical authority, which Google officially recognised in May 2023. However, adapt the approach for AI search: lead sections with extractable answers, use standalone sections that make sense without context (the "Taco Bell Test"), and implement comprehensive schema markup.

Sources

• SparkToro/Datos — Zero-Click Search Study (July 2024)
• Gartner — Search Engine Volume Predictions (February 2024)
• Google Search Central Blog — E-E-A-T Guidelines Update (December 2022)
• Google Search Central Blog — Helpful Content Update (August 2022)
• Seer Interactive — AI Overviews CTR Impact Study (September 2025)
• NPR/Edison Research — Smart Speaker Ownership Study (2022)
• DemandSage — Voice Search Statistics (2024-2025)
• Search Engine Journal — Voice Search Analysis
• Search Engine Land — Topic Clusters and SEO (2025)
• Backlinko — Search Ranking Factors Study (11.8 million results)
• Ahrefs — Link-Ranking Correlation Study (1 million keywords)
• IAB UK — AI Usage Statistics (October 2025)
• OpenAI — ChatGPT Usage Statistics (2025)