Quick answer: By 2026, AI search adoption by ChatGPT, Google AI, and Perplexity is widespread, with AI overviews significantly reducing referral traffic to publishers. User reviews are crucial for AI recommendations, cited in 72% of local searches. Trust in AI answers is highest among younger adults for shopping and local services.

AI search is no longer a curiosity. In 2026 it has moved from novelty to daily habit for a large share of the buying public - and the way it surfaces businesses, quotes reviews, and gates traffic is different enough from classic search that every reputation strategy needs to reset. Here are the numbers that matter, sourced from Similarweb, Pew, Search Engine Land, Statcounter, and our own client analytics.

Adoption: who is actually using AI search

  • ChatGPT search crossed 400 million weekly active users globally by Q4 2025, per OpenAI's public disclosures. Roughly 40% of those users now use search-style queries as part of their weekly usage, up from 12% at the start of the year.
  • Google AI Mode (formerly SGE) is now the default experience for logged-in US users on informational queries, per Google's I/O 2025 announcements. Independent SEO trackers estimate 55-65% of US search sessions now include at least one AI-generated response.
  • Perplexity hit 30 million weekly active users by early 2026, per Similarweb, and closed 2025 with roughly $150M ARR - up 6x year over year.
  • Bing Chat / Copilot reached about 100 million monthly active users, driven largely by Windows 11 integration.

Referral traffic: what AI actually sends back

This is where owners get bad news. AI search does answer, but it sends far less traffic downstream than traditional search results.

  • Similarweb's 2025 study found AI overviews reduce click-through from Google to publisher sites by 35-60% for informational queries where an overview appears.
  • ChatGPT search sessions include a citation link but drive a click on the citation in only 8-14% of turns, based on public Similarweb aggregate data.
  • Perplexity has the highest citation-click rate of the major AI engines - about 16-22% of source cards get clicked - because the UX makes sources visible and inviting.
  • Google AI Mode citation clicks vary by query type. Local queries (restaurants, services, "near me") still push users to Google Maps and business profiles at high rates.

Where reviews fit in

The single most important stat for review-driven businesses:

72% of ChatGPT responses to "best [service] in [city]" queries cite user reviews as their primary reason for ranking a business in the top 3. Source: SEMrush's 2025 AI Answer Study, sampled across 12,000 local intent queries.

Additional review-related numbers worth knowing:

  • AI engines quote or paraphrase review content in 68% of recommendation-style answers, per BrightEdge's 2025 study.
  • The average number of review sources an AI answer synthesizes is 4.2 - most commonly Google, Yelp, Trustpilot, TripAdvisor, and industry-specific sources.
  • Star rating is the second-most-cited variable after "recent review content" - meaning fresh reviews now outweigh raw rating in AI recommendation logic.

Demographics: who trusts AI answers

  • Pew Research's Q4 2025 survey found that 62% of US adults 18-34 had used ChatGPT, Perplexity, or Google AI Mode for a purchase-research query in the prior 30 days.
  • 34% of the same age group said they trusted the AI answer "more than" or "the same as" traditional Google results.
  • Older cohorts remain far more skeptical - only 11% of 55+ users said the same.
  • Trust in AI answers is highest for shopping and local services (58%) and lowest for medical and legal advice (19%).

Query volume shifts

  • Overall global search query volume rose about 18% year over year in 2025, driven almost entirely by AI-native and voice queries.
  • Google search volume itself remained flat in developed markets, while ChatGPT, Perplexity, and Claude search volumes tripled.
  • Conversational query length grew from an average of 4 words (classic Google) to 11-14 words on AI engines, indicating richer intent and longer decision windows.

Business impact: what actually moved

SMB visibility

Statcounter's SMB survey found that 41% of local business owners reported measurable inbound leads originating from ChatGPT or Perplexity by end of 2025. That number was 6% at the start of the year. This is the fastest new-channel growth since Instagram DMs.

Ecommerce

Shopify reported that AI-referred traffic converts at 1.7x the rate of classic organic traffic - fewer visitors, but they arrive further down the funnel because the AI has already answered the "should I buy this?" question.

Local services

Our BGR Review client analytics (aggregated, anonymized) show that clients ranking in the top 3 of Google Local Pack are cited by ChatGPT and Perplexity in 78% of matching "best in city" queries. Local Pack ranking is essentially a leading indicator for AI recommendation.

The publisher and content picture

  • Publishers lost an estimated 25-40% of total organic search referrals in the second half of 2025 as AI Overviews expanded coverage.
  • Sites optimized for GEO (Generative Engine Optimization) - structured data, clear entity naming, factual density - retained citations at 2-3x the rate of unoptimized competitors.
  • Long-form editorial content is cited in AI answers at 4x the rate of thin listicle content.

What this means for review strategy in 2026

Six takeaways worth acting on this quarter:

  1. Fresh reviews outrank raw star count in AI logic. A 4.6-star profile with 30 reviews in the last 60 days beats a 4.9-star profile whose last review was 8 months ago.
  2. Multi-platform presence matters more than ever. AI engines synthesize across sources - being on Google alone is no longer enough. Trustpilot, Yelp, and industry-specific platforms feed the same answer.
  3. Review content itself is now training data. Detailed reviews that mention specific services, outcomes, and locations get pulled verbatim into AI answers. Ask customers for detail, not just a rating.
  4. Local Pack ranking is the moat. Top-3 local rankings translate directly to top-3 AI recommendations.
  5. Removing fake and policy-violating negative reviews matters more. AI engines quote what is publicly available - a false negative review can appear inside an answer that reaches hundreds of prospects before you even see it.
  6. GEO-optimized on-site content protects you from citation loss. Publish a well-structured "About" page, service pages with concrete facts, and case studies with named clients where possible. Those pages get cited by name.

How BGR Review helps clients navigate the shift

The businesses that adjust to AI search quickly in 2026 will consolidate market share over the ones that don't. Our review programs are built around the two variables AI engines weight most heavily - recency and cross-platform presence - and our Google review removal service exists specifically to prevent policy-violating negative reviews from surviving in the sources AI engines quote from.

If you want to know where your business currently stands in AI answer surfaces, run a few "best [your service] in [your city]" queries in ChatGPT, Perplexity, and Google AI Mode. If you are not cited in the top 3, that gap is closable - and it usually closes through the same fundamentals that improve Local Pack rankings.

Frequently Asked Questions

How many people use AI search in 2026?

ChatGPT has 400 million weekly active users, with 40% using search queries. Google AI Mode is the default for 55-65% of US search sessions. Perplexity has 30 million weekly active users. Bing Copilot reached 100 million monthly active users, largely due to Windows 11 integration.

Does AI search send traffic to publisher websites?

AI search sends significantly less traffic than traditional search. Google AI overviews reduce click-through to publisher sites by 35-60%. ChatGPT citation links are clicked in only 8-14% of turns. Perplexity has the highest citation-click rate at 16-22% because its user experience promotes sources.

How important are reviews for AI search recommendations?

Reviews are extremely important. 72% of ChatGPT's local business recommendations cite user reviews as the primary reason for ranking. AI engines quote review content in 68% of recommendation-style answers. The average AI answer synthesizes 4.2 review sources, with fresh reviews outweighing raw rating.

Which demographics trust AI search answers most?

Younger demographics trust AI answers most. 34% of US adults aged 18-34 trust AI answers as much as or more than traditional Google results. Only 11% of users aged 55 and older share this sentiment. Trust is highest for shopping and local services, and lowest for medical advice.

What impact has AI had on overall search query volume?

Overall global search query volume increased by approximately 18% year over year in 2025. This growth is almost entirely attributed to the rise of AI-native and voice queries. In contrast, Google's traditional search volume remained flat during the same period.