01The search behavior shift
Discovery is moving from "ten blue links" to AI-generated answers — fast.
Gartner forecasts traditional search engine volume will fall 25% by 2026 as AI chatbots and virtual agents absorb queries.
Google AI Overviews now appear on ~48% of all search queries (Mar 2026), up from 34.5% in Dec 2025.
Question-style searches (who/what/why) trigger an AI summary about 60% of the time; one- or two-word queries only ~8%.
ChatGPT reached ~900M weekly active users (Feb 2026), more than double the 400M of Feb 2025.
02Clicks & traffic impact
When the answer is on the results page, fewer people click through — unless you're cited inside it.
Users click a result 8% of the time when an AI Overview shows, vs 15% without — a 46.7% relative decline across ~68,000 queries.
A field study found AI Overviews cut organic click-through ~38% for affected queries.
Brands cited inside an AI Overview earn ~35% more organic and ~91% more paid clicks than the result in position one below it.
Adobe measured generative-AI referral traffic to US retail sites up 1,200% YoY (and +4,700% YoY at its July 2025 peak).
03AI traffic quality & conversion
Lower in volume, higher in intent — AI-referred visitors convert and engage more.
ChatGPT referral traffic converts at 7.1% — second only to paid search (7.8%) and ahead of organic, direct, social and email.
AI-referred shoppers converted 31% more often than non-AI sources over the 2025 holiday season, with 32% longer visits.
AI-search referral traffic converts at 4.4× to 23× the rate of organic search visitors.
By engine, conversion rates ran ChatGPT 15.9%, Perplexity 10.5%, Claude 5%, Gemini 3%.
04B2B buyers have already moved
For B2B, AI answer engines are now a primary research surface — before a vendor's own site.
89% of B2B buyers have adopted generative AI, naming it a top source of self-guided research across every buying phase.
Generative AI has overtaken traditional search for ~25% of B2B buyers as their research starting point.
A 2026 survey found 96% of B2B companies are effectively invisible in AI discovery.
05Citations & visibility: the engines disagree
Each answer engine draws from a different source pool — winning one says little about the others. This is the core AEO problem.
Across 680M citations, only 11% of domains are cited by both ChatGPT and Perplexity (cross-engine overlap 6–16.4%).
Brand-citation rates vary 46× across engines — ChatGPT cited brands 0.59% vs Perplexity 13.05% (Grok ~27%), over 34,234 responses.
ChatGPT citations show ~87% overlap with Bing's top-10 rankings — its sourcing leans heavily on Bing.
The AEO/GEO tooling category has crossed $200M+ in disclosed funding as of early 2026.
06The persona dimension: one ranking per customer type
It isn't only that engines disagree — the same engine returns different brands per buyer persona. Average that away and you optimize for a customer who doesn't exist.
Across ~2,000 model runs, mid-market brands saw up to 75% of their AI recommendations replaced when the buyer persona changed — same query, different stated customer type. Established brands stayed stable.
"A brand doesn't have one AI ranking — it has one per customer type." Generic prompt tracking reports an averaged, non-existent buyer; AI visibility has to be read per persona to be real.