We analyzed over 117,432 inbound leads across multiple B2B brands to understand how AI search traffic compares to traditional search and other sources in terms of pipeline quality and downstream impact.
As more buyers turn to AI search and LLMs to ask product and vendor-related questions, we set out to answer the most critical question:
This question led us to conduct one of the largest B2B focused AEO studies in the AI search space.
The study analyzed traffic from AI platforms (ChatGPT, Perplexity, and Gemini) compared to traditional search engines (Google and Bing as well as other traffic sources including paid, direct and referral traffic.
The findings aren’t just interesting, they’re a wake-up call for growth leaders, SEO teams, and content marketers clinging to legacy strategies and models of attribution and optimization.
We analyzed data from B2B SaaS and professional services companies, across the U.S., with ARR ranging from $1M to $108M.
Over 117,432 inbound leads were reviewed from October 2024 through April 2025. All data used last-click attribution and tracked full-funnel performance:
Traffic → Leads → MQLs → SQLs → Opportunities → Closed Won
We segmented traffic into three categories:
Despite the lower volume, AI search traffic converts into pipeline at significantly higher rates than traditional search.
That’s a 56.3% higher close rate from leads that originated in AI search agents compared to Google or Bing.
Among all platforms analyzed, ChatGPT is the most efficient B2B traffic source that exists today with nearly 2X the lead-to-close rate of traditional search engines.
Marketers are still spending billions optimizing for Google, while ignoring the fact that discovery behavior is shifting. AI-native agents like ChatGPT and Perplexity are becoming front doors to product research, vendor comparisons, and commercial queries.
Here’s what we observed:
We identified four key factors that drive the increased funnel conversion rates from AI search traffic.
Google continues to dominate B2B website traffic, despite the rapid growth in AI search platforms. ChatGPT emerged as the second-largest traffic source, surpassing Bing and other traditional search engines. Among AI platforms, ChatGPT leads decisively with 6.5x more traffic than Perplexity and 17.4x more than Gemini.
This is no longer news at this point; If you’re not optimizing for AI search, you’re missing out on where the highest-quality organic demand is forming. Here’s how to start:
Set up referrer tracking for traffic from ChatGPT browser plugins, Perplexity, and other assistants. Use UTM parameters or real-time traffic enrichment tools to segment that data.
Use an AEO platforms, such as Goodie, to observe and monitor where and how your brand is being mentioned in answers. Is your content being cited? Are you showing up in comparisons? Are competitors more or less visible than your brand? From this information you can develop a strategy that addresses topical/perception gaps to improve visibility.
LLMs value clarity, citations, structure, and verifiable claims. Don’t just write for humans; write for retrieval. Write to be summarized.
Track AI search-driven leads as their own cohort. Measure their MQL-to-Closed rates separately. Let the data speak for itself.
To help B2B marketers understand what “good” looks like for AI search traffic, we’ve broken down the key performance metrics by platform. These benchmarks can act as a guide to evaluate your own AI search performance.
The reality is search has changed. Buyer behavior has changed and the resulting funnel has changed. The highest-quality leads now come from ChatGPT, not Google. If your strategy hasn’t evolved, you’re not just behind in fact, you’re optimizing for the wrong benchmarks entirely. This study highlights the opportunity for brands that get in early, measure the right things, and optimize where the next conversion happens, not just where the last one did.
---
This study analyzed the full-funnel performance of 117,432 inbound leads from U.S.-based B2B SaaS and professional services companies between October 1, 2024 and April 30, 2025.
The goal: to determine whether AI search traffic drives better pipeline outcomes than traditional search or other channels. Claude Sonnet 3.7 was used to help analyze and normalize the data.
We standardized funnel stages across all companies:
Traffic was grouped into: