LLM leads grew 29% from January to March. February dipped 9% before March surged 42%. The number of accounts receiving at least one LLM lead also increased, from 371 in January to 480 in March, a 29% increase in the breadth of LLM-referred traffic.

As a share of organic, LLM traffic moved from 1.65% in January to 1.93% in March. That is a small number. It is also the kind of number that compounds. If the January-to-March growth rate holds, LLM traffic will exceed 3% of organic by Q4 2026.

ChatGPT dominates with 83% of all LLM-referred leads. Google Gemini is a distant second at 8%, followed by Perplexity (4.3%), Microsoft Copilot (4.1%), and Claude (0.9%). This mirrors the broader consumer adoption curve. ChatGPT has the largest user base and the most direct integration with search-like behavior.

Google Gemini's 8% share is worth watching. As Google continues to integrate AI into search results (AI Overviews, Gemini-powered SGE), the line between "organic search" and "AI-referred traffic" will blur. Some of what we currently classify as Google Organic may already be influenced by AI-generated results.

LLM leads convert at a lower rate than organic. The book rate for LLM-referred leads is 33.6% compared to 43.2% for organic searchma 10-point gap. The paying customer rate follows the same pattern: 15.2% for LLM vs. 23.0% for organic.

This makes sense. A searcher who types "plumber near me" into Google has clear service intent. A user asking ChatGPT "who's a good plumber in Dallas" may be earlier in the research process, comparing options, or not yet ready to book.

That said, LLM leads are still converting. A 33.6% book rate and $1 million in closed revenue across one quarter is not noise. These are real customers finding contractors through a new channel.

One notable detail: Microsoft Copilot has the highest book rate at 41.1% and the highest paying customer rate at 18.9%, approaching organic-level quality, though on a small sample of 95 leads.

The biggest challenge with LLM benchmarks is that the data is not evenly distributed. The median contractor received 2 LLM leads in all of Q1. At the 90th percentile, that number jumps to 7 leads. The top account received 37 leads.

For most contractors, LLM traffic isn't high volume yet based on this sample data. But for the top 10% of accounts, LLM traffic represents 7.8% or more of their organic volume. And at the very top, one contractor saw LLM-referred leads reach 23% of their organic search volume in Q1.

Why do some contractors get 20x more LLM traffic than others? We do not have a definitive answer, but the pattern suggests that contractors with strong branded presence — those who show up in local directories, review sites, and content that LLMs train on are more likely to be recommended by AI. It's also based on search demand from the LLMs themselves and it is possible some markets have a population who is more actively using LLMs than others.

LLM traffic to home services contractors is real, growing, and generating revenue. It is also small, 1.74% of organic, and heavily concentrated among a subset of contractors.

The honest framing is this: LLM traffic is growing but is still not yet a meaningful revenue generator on average. The trajectory,29% growth in a single quarter, expanding from 371 to 480 accounts receiving leads, reinforces our instincts about the rise of AI and how it will impact home service businesses from a customer generation perspective.

We will update this benchmark quarterly as the dataset grows. Sometimes new technology adoption happens slowly then suddenly, so you'll want to continue to watch this consumer behavior trend closely!