Most website analytics tools tell you how many people visited your site. They rarely tell you who. When that traffic arrives from an AI platform like ChatGPT or Perplexity, the gap between "session recorded" and "lead identified" is where pipeline goes to die. This article walks through five concrete steps to close that gap: tracking which AI platforms are sending you visitors, identifying the companies and individuals behind anonymous sessions, and handing those names directly to your sales team.

TL;DR

• AI referral traffic is growing fast, but standard analytics platforms do not identify the people behind those visits [wearetg.com]

• Reverse IP lookup and identity resolution tools can de-anonymize a meaningful share of your anonymous visitors [marketbetter.ai]

• The process works in five steps: tag your traffic sources, segment AI referrals, resolve company identity, surface individual contacts, and route to sales

• LLM visibility tracking is a prerequisite: you cannot convert traffic you cannot see

• Real-world examples show this approach surfaces leads that would otherwise be permanently lost

About the Author: Simaia is an agentic marketing team that helps B2B companies get found in AI search results across ChatGPT, Gemini, Claude, Perplexity, and Google AI Overview, then identifies the visitors those results send. Simaia has de-anonymized inbound leads for clients across APAC, including surfacing a high-value healthcare buyer for an Australian SaaS company that had zero AI search visibility just 2.5 months earlier.

Why Does AI Referral Traffic Need Special Treatment?

AI referral traffic behaves differently from every other channel you are currently tracking. When a buyer asks ChatGPT "which B2B logistics software is best for Southeast Asia" and your brand appears in the answer, the resulting click arrives with minimal referral data. Standard UTM parameters are absent. The session looks nearly identical to direct traffic. Without deliberate tagging and segmentation, you will never know the visit happened [wearetg.com].

This is not a minor data-quality issue. It is a structural blind spot. As more B2B buyers use AI platforms as their primary research tool, the share of "invisible" high-intent traffic on your site grows with it. Fixing the problem starts with making AI traffic visible before you can even think about identifying who visits your website.

Step 1: How Do You Tag and Track AI Referral Sources?

The first step is building a tracking layer specifically for AI platforms. Out of the box, Google Analytics 4 does not classify ChatGPT or Perplexity as distinct channels. You need to create a custom channel group using regex patterns that capture traffic from these sources [dkm.agency].

Platforms to include in your regex rules:

• ChatGPT (chat.openai.com)

• Perplexity (perplexity.ai)

• Google AI Overview (google.com with AI referral signals)

• Gemini (gemini.google.com)

• Claude (claude.ai)

Once this channel group is live, every session originating from an AI platform is flagged separately in your reports. This is the foundation for everything that follows. Without it, LLM visibility tracking is essentially guesswork [dkm.agency].

Step 2: How Do You Segment AI Visitors From the Rest of Your Traffic?

Building on the tagging layer above, the harder question is separating AI-referred human visitors from AI bot traffic. Both show up in your logs, and conflating them produces inflated session counts and misleading conversion data [wearetg.com].

Two categories to separate:

• AI bot visits: Crawlers from LLM training runs and index bots (e.g., GPTBot, ClaudeBot). These inflate raw traffic figures but represent no human buyer.

• AI referral visits: A human who read an AI-generated answer and clicked through to your site. These are your actual prospects.

To segment them, filter by user-agent strings to exclude known bot signatures. What remains in your AI referral channel group represents real people who found you through an AI answer. This is the audience you want to identify anonymous website visitors within.

Step 3: How Does Reverse IP Lookup Identify Companies Behind Anonymous Sessions?

This is where anonymous visitor identification becomes actionable. When a visitor lands on your site, their device sends a request from a registered IP address. Many corporate IP ranges are mapped to the companies that own them. A reverse IP lookup (rDNS) cross-references that address against a database of these mappings and returns the company name, industry, size, and location [marketbetter.ai] [hockeystack.com].

What reverse IP lookup typically surfaces:

• Company name and domain

• Industry vertical

• Employee count range

• Geographic location

What it does not reliably surface:

• The individual's name

• Their direct email or phone number

• Their LinkedIn profile

That individual layer requires the next step. It is worth noting that reverse IP lookup works best for corporate network traffic. Visitors on residential ISPs or VPNs are harder to resolve, which is why identity resolution tools layer additional signals on top of the IP data [vison.ai].

Step 4: How Do You Surface Individual Contact Details From Anonymous Visits?

Stepping back from the technical detail, a separate concern is moving from "a company visited" to "a specific person visited." This is where identity resolution tools and first-party data enrichment come in [factors.ai] [leadpipe.com].

The layered approach:

When these signals are combined, a meaningful share of anonymous visitors can be resolved to a named individual [vison.ai]. For B2B traffic specifically, where a single visitor might represent a six-figure deal, even a modest identification rate produces significant pipeline value.

Real example: For an Australian Healthcare SaaS client, Simaia's identification layer surfaced a major healthcare organization's inbound visit, including the individual contact and their LinkedIn profile, after that visitor arrived via an AI referral. The sales team was able to follow up directly on a lead that would have otherwise vanished as an anonymous session.

Step 5: How Do You Route Identified Leads to Your Sales Team?

Identifying a lead is only half the job. The information needs to reach the right salesperson at the right moment, before the intent signal goes cold [aisdr.com].

A practical routing workflow:

1. Visitor arrives from an AI referral source (flagged by your channel group)

2. Reverse IP and identity resolution run automatically in the background

3. Company and individual details are resolved and logged

4. A notification is triggered to the relevant salesperson (by industry, territory, or deal size)

5. The salesperson reaches out within the same business day

Speed matters here. A buyer who just found you in a ChatGPT answer is in active research mode. A same-day outreach is qualitatively different from a follow-up three days later [aisdr.com].

Frequently Asked Questions

Can I identify every anonymous visitor to my website?
No. Identification rates vary based on IP type, browser settings, and whether the visitor has a known digital footprint. Corporate visitors on company networks are the most resolvable [marketbetter.ai].

Is identifying website visitors legally compliant?
Company-level identification via reverse IP is generally considered lawful. Individual-level identification should align with your privacy policy and applicable regulations in your market [vison.ai].

What percentage of AI referral visitors can typically be identified?
Identification rates vary by industry and traffic quality. B2B traffic from corporate IP ranges has higher resolution rates than consumer traffic.

How is this different from standard Google Analytics?
GA4 tracks sessions and behaviour but does not identify the company or individual behind a visit. Visitor identification tools add a resolution layer on top of your existing analytics [factors.ai].

Do I need separate tools for AI traffic tracking and visitor identification?
Often yes. AI traffic tracking requires custom channel configuration in your analytics platform [dkm.agency], while visitor identification requires a separate identity resolution service [leadpipe.com].

How long does setup take?
Custom channel groups in GA4 can be configured in under an hour. Identity resolution tools typically require a tracking pixel and take one to two days to begin producing data.

What if a visitor uses a VPN or residential internet?
These sessions are harder to resolve at the company level. Individual-level identification through cookie matching or form enrichment becomes more important for these cases [hockeystack.com].

About Simaia

Simaia is an agentic marketing team that replaces the need to hire a marketing manager, content writer, SEO consultant, and lead intelligence vendor separately. Built for B2B companies across APAC, Simaia handles the full AI visibility workflow: auditing where your brand appears across ChatGPT, Gemini, Claude, Perplexity, and Google AI Overview; writing and placing content that gets cited by those models; and identifying the companies and individuals who arrive from AI referrals. For a global textile manufacturer, this approach grew inbound leads from one every two months to five per month within two months. For a healthcare SaaS in Australia, it took AI search visibility from 0% to 45% of niche traffic in 2.5 months.

If you want to see exactly who is visiting your website from AI referral traffic and turn those visits into named, contactable leads, visit simaia.co to learn more or get in touch with the team.