• by Laurie Sullivan , Staff Writer, September 30, 2026

AI search has become a measurable channel for discovery, decision-making and commercial growth, as the pace of growth continues to surge.

Sessions have more than doubled in the past year since July 2026, tripled during the most recent quarter, and grew nearly fivefold during the past 13 months, according to data released this month.

Based on 1.5 million AI search sessions across 175 websites and 20 industries over 13 months, data from the “State of AI Search 2026” report published by IDHL, a digital agency that specializes in performance marketing, shows cross-media views yet of how AI powers brand discovery.

The findings reveal AI search grew 392% year-on-year (YoY), and revenue from those using the sites increased 312%.

Transactions rose faster than traffic, up 553% YoY, which suggests AI search is becoming a “commercial opportunity for brands.” In fact, 21,141 total transactions were generated through AI during the past year. The report suggests AI-powered discovery influences higher-intent customer journeys, not just top-of-funnel research, according to the data.

Findings also show that AI search now outpaces every major acquisition channel, growing roughly 14 times faster than organic search, which increased 28% during the past year, compared with 19% growth for paid search, and 25% for direct traffic.

Brands can no longer build visibility around one platform, particularly as Google and Microsoft push AI deeper into search, devices and workplace tools at a head-spinning pace.

AI search is starting to follow seasonal demand, and most importantly, the data suggests AI search is beginning to follow the same seasonal patterns, category spikes and shifts in consumer demand that have long shaped traditional search.

Marketers can use the seasonality to plan content and optimize activity ahead of category peaks, Chris Shelbourn, organic search director at IDHL, told MediaPost.

AI searches for "gifts and travel peak in December,” Shelbourn wrote in an email. “Across the board, most categories peak in May and June, which I predict is due to a wider industry shift reported with more clickable links appearing in ChatGPT.”

He shared a chart showing peak months for major retail sectors and their year-over-year growth.

Trade and Do It Yourself sectors grew 653% from July 2025 through July 2026, with search peaking in July, for example. Home & Garden at 588% growth, respectively, also peaks in July.

Entertainment at 419% growth peaks in May, and Beauty at 418% growth peaks in June.

The State of AI Search 2026 shows that marketers and advertisers need to think ahead. AI visibility doesn’t begin when someone asks ChatGPT or Gemini a question. It starts months earlier with the content and recommendations that shape how a brand is understood online, according to the report.

OpenAI ChatGPT leads overall, but Google Gemini and Anthropic Claude are gaining. While ChatGPT remains the dominant source of AI search traffic at 91%, newer platforms, including Gemini, Claude, Copilot and OpenAI Search are gaining momentum, creating a more competitive landscape for brands seeking visibility within AI-generated answers.

When asked about the obsession with pulling information from Reddit into AI search, Shelbourn explained that “AI engines tend to trust the kind of content Reddit is full of, which is real people sharing real experiences."

He added that this has made it a " frequently cited third-party source in AI search answers across a wide range of industries and topics and, inevitably, a target for digital marketers.”

One of the more interesting stats in the report is that real opinions from real people now carry more weight than ever. Last year alone, users chose a Reddit result 23 billion times within Google searches, actively seeking answers that cut through low-quality content.

He also explained that while participating in relevant Reddit conversations have become a viable tactic based on generative engine optimization, "it is perhaps not the 'silver bullet' some commentary suggests," because large language models "regularly cite dozens of other third-party sources, which can vary substantially by industry or topic, not just Reddit."