Targetlytics.AI
Identify pillar

The companies AI sent you, by name

Most AI-sourced traffic never identifies itself. Visitor Identification resolves those sessions to real organisations, links each one to the prompt that produced it, and scores how close they are to buying.

An account, not a line in an analytics report

Each identified visit becomes a record you can act on:

The company

Name, domain, and firmographics from licensed business data.

The origin

The prompt you were cited on, the citation that carried it, and the assistant that produced the answer.

The journey

Pages viewed, order, and time on each.

Intent

A score built from what they did, weighted toward the pages that indicate buying: pricing, contact, product detail.

Persona

Matched against the personas that actually close in your CRM, not a generic segment.

Return behaviour

Repeat visits recognised and folded into the same account, which is where the signal gets strong.

Signals, in order of reliability

AI-sourced traffic is difficult to attribute because assistants strip or flatten the usual signals, and much of it lands in analytics as direct. We read several signals and fall back through them rather than depending on any one.

  1. 1
    Referrer
    Where it survives, it's decisive.
  2. 2
    Campaign parameters
    Where present.
  3. 3
    Landing URL matched against citation forensics
    This is the one that does the heavy lifting. Because we know which of your pages are cited for which prompts, an arrival on a cited page is strong evidence of both the source and the question behind it.
  4. 4
    Session and device signals
    Used to hold a session together and recognise a return.

Step 3 is why this works when referrer data doesn't survive — and it is only possible because the prompt tracking and citation forensics already exist underneath. A standalone visitor-identification tool has no equivalent.

When it doesn't work

  • Company resolution is worldwide. Person-level resolution is US only — never for EEA or UK visitors.
  • VPNs, consumer ISPs and mobile networks often resolve to nothing usable.
  • Very small companies resolve less reliably than large ones.
  • A visit that can't be resolved with confidence is reported as unidentified. We don't fill gaps with guesses.
  • Identification improves on return visits, because there's more behaviour to work with.

One script, then two connections

  1. Install the tracking script — one tag, no build step
  2. Connect Google Search Console and GA4
  3. Connect your CRM so personas come from deals that actually closed
  4. Set your attribution inputs — CAC, lead-to-conversion rate, and your sales parameters

The script sets a first-party cookie (_tl_vid) and a matching local storage entry so a returning visitor is recognisable. In the EEA and UK this generally requires consent under ePrivacy. You're the data controller for your site — set window.targetlyticsConsent = 'denied' before the script, or call targetlytics.revokeConsent(), to keep attribution running without the cookie or any visitor record.

Related reading

GDPR-compliant visitor identification
Company-level identification an EU DPO will approve.
RB2B alternative for Europe
Person-level vs company-level — and where each one fits.

See who AI is sending you

Visitor Identification ships with the Identify pillar on Ultimate.

Compare plans