Market data and deal flow · No. 06 / 13
How fast is AI-led discovery growing—and what do the volumes mean?
Last updated
Discovery growth supports investigation, not a universal spending prescription. The available datasets measure different populations and outcomes.
| Indicator | Figure and observation period | Source date | Buyer-side interpretation |
|---|---|---|---|
| Global generative-AI web usage | Similarweb reports 9.5 billion average monthly visits, 70% higher year on year, for June 2025–May 2026 | Statistics page accessed 16 October 2026 | Estimated visits across generative-AI platforms—not US buying queries or unique people. (aisearch.similarweb.com) |
| AI referrals to US retail sites | 393% year-on-year growth, January–March 2026 | Adobe, 16 April 2026 | Strong growth in Adobe’s retail dataset; does not establish AI’s absolute traffic share for a particular SaaS business. (business.adobe.com) |
| Conversion of AI-referred US retail visits | 42% higher than non-AI traffic in March 2026, compared with 38% lower in March 2025 | Adobe, 16 April 2026 | Relative conversion differences, not percentage-point changes or proof that GEO work caused additional sales. (business.adobe.com) |
| B2B buyer use of LLMs | 94% in research covering nearly 4,000 buyers, associated with the 2025 Buyer Experience Report | 6sense guide, 25 November 2025 | Global, supplier-produced survey evidence—not a representative estimate for every US SaaS category. (6sense.com) |
| Clicking when Google shows an AI summary | Traditional-result clicks on 8% of visits with a summary versus 15% without; summary-source clicks on 1% of visits with a summary | Pew, published 22 July 2025, based on 900 US adults’ March 2025 browsing | Historical observational evidence; not a current click-rate forecast or causal experiment. (pewresearch.org) |
For a B2B growth leader, the practical question is not whether AI usage is large. It is whether your buyers use it for the questions that determine qualification, comparison and shortlisting. Ask customers, inspect attributable referrals and maintain a separate record of recommendation visibility. Do not substitute retail conversion benchmarks for a SaaS business case.
Sources cited on this page
- aisearch.similarweb.com aisearch.similarweb.com/blog/gen-ai-stats/?utm_source=openai
- business.adobe.com business.adobe.com/blog/ai-traffic-surge-retail-sites-not-machine-readable?trk=a
- 6sense.com 6sense.com/guides/how-genai-and-llms-are-changing-b2b-buyer-research-and-how-to-
- pewresearch.org www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click