Your Buyers Are Building Shortlists Inside LLMs. You're Not on Them.
The B2B buying journey has fundamentally shifted. Buyers no longer start with a Google search — they start with a conversation. They ask ChatGPT, Claude, Gemini, or Copilot to recommend vendors, compare features, and build a shortlist. By the time they visit your website, the decision is nearly made. And you have no idea they were ever there.
The LLM Dark Funnel: Shortlists Built Where You Can't See
A procurement director types 'recommend the top 5 visitor identification tools for B2B SaaS' into Claude. The LLM synthesises data from review sites, documentation, community discussions, and analyst reports to produce a ranked shortlist with pros and cons. The director's buying decision is 70% made — and your CRM shows zero activity. This is the LLM dark funnel: the most influential stage of the buying journey now happens inside a conversational AI where you have zero visibility, zero attribution, and zero ability to influence the outcome.
LLM Click-Throughs Arrive as 'Direct' Traffic — Your Analytics Is Blind
When a buyer reads an LLM-generated vendor comparison and clicks a link to your website, the HTTP referrer header is typically stripped or obscured. GA4 records the visit as 'Direct / None' — the same bucket as someone who typed your URL from memory. You're receiving ultra-high-intent visitors who were just recommended your product by an AI assistant, but your analytics treats them identically to a bookmark click. Budget and attribution decisions are made completely blind to your fastest-growing inbound channel.
LLM-Educated Buyers Move Faster Than Your Sales Process
Traditional B2B buyers take weeks to progress through awareness, consideration, and decision stages. LLM-educated buyers compress this entire journey into a single conversation. A buying committee member can research your category, get vendor recommendations, read synthesised feature comparisons, and form an opinion — all in 15 minutes inside ChatGPT. By the time they visit your website, they're in final validation mode. If you wait for them to fill a form (only 2-3% ever will), they've already scheduled a demo with the competitor the LLM ranked higher.
85% of Buyers Trust LLM Recommendations — But You Can't Measure the Impact
Research shows that 85% of B2B buyers think more highly of a vendor when an AI chatbot mentions them in a recommendation, and many change their final vendor choice based on LLM-surfaced information. This means LLMs are actively shaping your pipeline — for better or worse — and you have no measurement framework to quantify it. Your marketing team can't prove LLM-driven impact because the traffic arrives unlabelled, the visitors stay anonymous, and the influence happens in a closed conversational environment.
Turn Every LLM Click-Through Into a Named, Qualified Lead
Kwin identifies the B2B decision-makers who arrive at your site after researching you inside any Large Language Model — ChatGPT, Claude, Gemini, Copilot, or Perplexity. No forms. No pop-ups. No dependence on referrer data.
Install Kwin on Your Website
Add Kwin's lightweight first-party pixel to your website — directly or via Google Tag Manager. Kwin identifies visitors independently of how your analytics classifies the traffic source. Whether a buyer clicks a link inside a ChatGPT conversation, copies a URL from Claude into their browser, or follows a Copilot citation, Kwin resolves their identity the moment they land. Setup takes under 5 minutes with zero IT involvement.
Identify LLM-Referred Visitors at the Person Level
The moment an LLM-referred visitor lands on your site, Kwin's proprietary identity graph resolves them — full name, verified business email, job title, company, LinkedIn profile, and direct phone number. This works whether they arrived via a direct LLM link (which strips referrer data) or a copy-pasted URL (which appears as 'Direct' in GA4). You'll know that the anonymous visitor who went straight to your pricing page is actually John Doe, Head of Revenue at a 400-person fintech — who just asked Claude to compare your product against three competitors.
AI Detects the 'LLM-Educated Buyer' Pattern
LLM-referred visitors behave differently from every other traffic source. They skip your blog, ignore your top-of-funnel content, and navigate directly to decision-critical pages: pricing, security documentation, API references, integration guides, and customer case studies. Kwin's intent scoring engine detects this distinctive 'LLM-educated buyer' pattern and assigns elevated Purchase Intent Scores — flagging them as prospects who have already completed their research and are in final validation mode.
Automated Outreach That Matches Their Research Depth
Qualified LLM-referred visitors automatically receive personalised email sequences calibrated to their stage — no introductory pitch, no generic nurture track. Outreach jumps straight to competitive differentiators, customer proof points, and meeting invitations. Your sales team gets instant Slack alerts with full context: the visitor's identity, company, role, pages evaluated, and intent score. Engage the buyer while they're still validating — before they finalise the shortlist the LLM helped them build.
Ready to see who's clicking through from LLMs?
Start free. No credit card. No contracts. Then pay-as-you-go at just $0.50 per identified lead.
Own the Channel That's Reshaping B2B Buying.
LLM-referred visitors are pre-qualified by the AI itself. They've already read synthesised product comparisons, feature analyses, and competitive evaluations inside the LLM conversation. Studies show this traffic converts at 4-6x the rate of standard search traffic. Kwin identifies 60-70% of these visitors at the person level — turning your highest-converting channel into predictable pipeline.
ChatGPT, Claude, Gemini, Copilot, Perplexity, and every LLM that launches next — Kwin identifies visitors from all of them with a single pixel. Each model handles outbound links differently: some pass referrer headers, some strip them, some open links in embedded browsers. Kwin's identification is completely independent of these mechanics. One installation, universal coverage, future-proof.
For the first time, connect LLM-driven discovery to actual pipeline revenue. While you can't see what happens inside the LLM conversation itself, Kwin reveals the person who emerged from that conversation and visited your site. Track how many identified leads came from LLM-educated buying behaviour, which pages they evaluated, and how they convert — giving your marketing team hard data on LLM influence.
LLM-educated buyers compress weeks of research into minutes. Their shortlist is often finalised within hours of starting their AI-assisted research. Kwin delivers instant Slack alerts the moment a high-intent LLM-referred visitor lands on your site — so your team can engage while the buyer is still in active validation mode, not after they've already committed to a competitor's demo.
From LLM Conversation to Booked Meeting — Fully Automated
Frequently Asked Questions
LLMs are building your buyers' shortlists right now.
Start identifying who clicks through.
Turn every LLM-referred click into a qualified lead. Identify the decision-makers who research you inside ChatGPT, Claude, Gemini, Copilot, and Perplexity — then engage them before the shortlist is finalised. Zero risk. 5-minute setup.