---
title: "What Is Website Visitor Identification? 2026 Guide"
url: https://brandengineering.ai/blog/what-is-website-visitor-identification
description: "Website visitor identification turns anonymous traffic into named companies or people. How account-level and person-level identification work, real match rate caveats, privacy rules, and how GTM teams wire it in."
updated: 2026-11-19
---

# What Is Website Visitor Identification? 2026 Guide

By Chris Riley. Published 2026-11-19.

Website visitor identification is the process of figuring out who is behind anonymous traffic on your site, either the company (account-level) or the individual person (person-level), so sales and marketing can act on visits that never filled out a form. Account-level tools match a visitor's IP address to a company. Person-level tools match the visit against an identity graph to return a name, work email, and title.

For B2B teams, a pricing page visit is one of the strongest intent signals you get. Visitor identification is how GTM engineers turn that signal into a routed lead, a scored account, or a retargeting audience. This guide covers how both approaches work, what match rates really mean, the privacy rules, the main use cases, and how to wire it into your CRM.

*Disclosure: Brand Engineering is built by the team behind Cuppa and Bullseye. We include them where they fit and say where they don't.*

## Account-level vs person-level website visitor identification

| | Account-level (reverse IP) | Person-level |
| --- | --- | --- |
| Returns | Company name, industry, size, pages viewed | Name, work email, title, company, sometimes LinkedIn profile |
| How it matches | Visitor IP against a database of company IP ranges | Cookies, device IDs, and IP matched against an identity graph |
| Best for | ABM, account scoring, sales prioritization | Speed-to-lead, direct outreach, person-level retargeting |
| Weak spots | Remote workers, mobile networks, VPNs, small companies | Coverage varies by geography and vendor; accuracy varies by signal |
| Privacy posture | Lower risk, still processes personal data (the IP) | Higher scrutiny; vendors often limit to the US |
| Examples | Leadfeeder (Dealfront), HubSpot (Clearbit) | Bullseye, RB2B, Warmly |

### How account-level identification works

You install a script. On each visit it captures the IP address and looks it up against a database of IP ranges registered to or observed for companies. [Leadfeeder describes it](https://help.leadfeeder.com/en/articles/13922307-frequently-asked-questions-about-the-leadfeeder-tracking-script) as matching "each visit's IP address against company databases," then adding company name and industry. [HubSpot's Buyer Intent](https://knowledge.hubspot.com/data-management/buyer-intent-intent-signals-and-custom-signals-frequently-asked-questions) does the same and states plainly that it identifies "the company behind a visit, not the individual person."

The weaknesses are structural. HubSpot notes that IP matching is "probabilistic": shared offices, VPNs, and ISPs can make a company appear when the visit came from someone unrelated. It also says [visits from mobile networks can't be identified](https://knowledge.hubspot.com/prospects/prospects-frequently-asked-questions), and smaller companies often show up as their ISP. With remote work, a lot of B2B traffic comes from home broadband that doesn't resolve to an employer at all.

### How person-level identification works

Person-level vendors match the visit against an identity graph: a database linking browser cookies, device identifiers, and IP addresses to known people, licensed from publisher and data partners.

- [RB2B says it](https://www.rb2b.com/) matches visitors "against a fully-permissioned publisher network using a combination of 1st and 3rd party cookies, device IDs and IP addresses," with Demandbase reverse-IP data for company-level matches.
- [Warmly's documentation](https://help.warmly.ai/articles/5643444357-overview-of-warmly-s-data-data-quality) describes a waterfall: filter bots and VPNs, check cookies (most accurate), then device fingerprinting, then probabilistic IP matching, querying multiple data vendors and scoring confidence when providers agree.

The practical consequence: person-level identification is only as good as the vendor's graph and how strictly it filters low-confidence matches. Two tools on the same site will return different people and different rates.

## Match rates: what the numbers mean

Every vendor publishes a match rate. None are independently audited, and they measure different things. Here's what vendors themselves say, as of November 2026:

| Vendor | Vendor-reported rate | What it covers |
| --- | --- | --- |
| [RB2B](https://www.rb2b.com/) | Up to 70-80% on Pro+; ~30-40% on Starter and Pro; ~15-20% on Free | People and companies combined on paid plans; Free is company-level only. Person-level is US-only |
| [RB2B help center](https://support.rb2b.com/en/articles/8891294-what-is-rb2b-and-how-does-it-work) | ~40-45% person-level, plus ~30-35% company-level | "Qualified US-based traffic" |
| [Warmly product page](https://www.warmly.ai/p/inbound-agent) | ~15% person-level at 90%+ accuracy | Inbound Agent, multi-vendor waterfall |
| [Warmly knowledge base](https://help.warmly.ai/articles/5643444357-overview-of-warmly-s-data-data-quality) | ~60% company, ~45% contact | Higher with Warmly UTM and form traffic |
| [Bullseye](https://bullseye.so/?utm_source=brandengineering&utm_medium=referral&utm_campaign=blog-what-is-website-visitor-identification) | Up to 40% of anonymous visitors | Name, work email, title, company |

Notice that Warmly's own pages give ~15% and ~45% for person-level identification. That isn't necessarily a contradiction; the numbers likely use different accuracy thresholds and denominators. It's the point: always ask what the denominator is (all traffic, US traffic, "qualified" traffic after bot filtering) and what accuracy threshold sits behind the number.

How to evaluate honestly:

1. **Run two tools side by side** on the same pages for two to four weeks.
2. **Spot-check identified visitors** against known traffic: your own team, customers who later filled out a form, people who booked a demo.
3. **Measure ICP matches, not raw matches.** A 40% rate full of students and competitors is worth less than 15% of real buyers.
4. **Track downstream conversion:** meetings booked and pipeline created from identified visitors.

## Privacy and compliance

This is not legal advice. Talk to counsel before you deploy, especially for person-level identification or non-US traffic. That said, here's how the rules map to the two approaches.

**GDPR (EU and UK).** IP addresses can be personal data. The EU Court of Justice held in [Breyer (C-582/14)](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A62014CJ0582) that a dynamic IP address stored by a website operator can be personal data when the operator has legal means to identify the person with additional information. So even account-level identification needs a lawful basis. [Leadfeeder's own guidance](https://www.leadfeeder.com/guides/website-visitor-identification/) puts it this way: company-level identification is lower-risk than person-level tracking, but it still involves personal data. Separately, ePrivacy rules govern cookies, so you may need consent before non-essential cookies are set. Leadfeeder supports [IP anonymization](https://help.leadfeeder.com/en/articles/13922307-frequently-asked-questions-about-the-leadfeeder-tracking-script) and cookie-less operation for this reason.

**Person-level outside the US.** Most person-level vendors restrict it. [RB2B states](https://www.rb2b.com/gdpr) that person-level ID is US-only, using IP geofencing, and that its global company-level ID requires a consent management tool.

**CCPA/CPRA (California).** California gives consumers [the right to opt out of the sale or sharing of their personal information](https://oag.ca.gov/privacy/ccpa), including through a browser-level Global Privacy Control signal. If your setup counts as selling or sharing under the CCPA, which counsel should assess, you'll need a "Do Not Sell or Share" mechanism and should honor GPC.

Practical defaults that keep you on solid ground:

- Disclose visitor identification in your privacy policy
- Use a consent management platform and gate scripts on consent where required
- Honor opt-outs and GPC
- Sign a DPA with your vendor
- Don't send cold outreach that reveals more than the visitor would expect ("I saw you spent 4 minutes on our pricing page" reads as creepy)

## Use cases

- **Speed-to-lead.** A target-account VP hits the pricing page; the owning rep gets a Slack alert within minutes, not after a form fill that never comes.
- **Routing.** Identified visitors are assigned by territory, account ownership, or round-robin, using the same rules as inbound leads.
- **Scoring.** Page views on pricing, integrations, and comparison pages feed lead and account scores.
- **Retargeting.** Push named, high-intent visitors to ad platforms as custom audiences instead of relying on anonymous cookie pools.
- **ABM.** Account-level data shows which target accounts are warming up, and how many people from each are researching.
- **Content feedback.** See which accounts read which articles, so content teams learn what moves pipeline.

## How GTM engineers wire visitor identification into the stack

The tool is the easy part. The value is in what happens after a match. A typical setup built by a [GTM engineer](https://brandengineering.ai/blog/what-is-a-gtm-engineer):

1. **Capture.** One script on the site. Exclude internal IPs and known bots.
2. **Filter.** Drop non-ICP matches by firmographics: company size, industry, geography, title seniority.
3. **Enrich.** Fill gaps (phone, tech stack, funding) through an enrichment waterfall, often in Clay or a similar tool.
4. **Dedupe and upsert.** Match against existing CRM contacts and accounts by email and domain before creating anything.
5. **Score.** Weight by page intent (pricing beats blog), recency, and repeat visits. Write the score to a CRM field.
6. **Route.** Above threshold, assign to the account owner and alert in Slack. Below threshold, add to a nurture sequence or ad audience.
7. **Measure.** Tag everything with a source so you can report pipeline from identified visitors against cost.

This is exactly the kind of work that shows up in [GTM Engineer postings](https://brandengineering.ai/jobs/gtm-engineer) on our board: GTM Systems, Salesforce, Automation, and CRM are among the most common tags. More advanced teams hand steps 3 to 6 to an agent; see [what an agentic operator does](https://brandengineering.ai/blog/what-is-an-agentic-operator).

### Where Bullseye fits

[Bullseye](https://bullseye.so/?utm_source=brandengineering&utm_medium=referral&utm_campaign=blog-what-is-website-visitor-identification), which we build, is a person-level visitor identification tool. Per its site, it returns name, work email, job title, and company for up to 40% of anonymous visitors, routes them into HubSpot, Salesforce, Slack, Twilio, and Sheets (with webhooks and Zapier), includes a B2B contact database for finding lookalike buyers, and syncs visitors to Meta, Google, and ChatGPT Ads audiences. It fits teams that want identification, enrichment, routing, and retargeting in one place. If you only need account-level data for ABM, or most of your traffic is in the EU, an account-level tool like Leadfeeder is the simpler fit.

## Alternatives

- **[RB2B](https://www.rb2b.com/):** Person-level identification of US visitors with LinkedIn profiles and business emails pushed to Slack, Teams, CRMs, and Clay, plus a free company-level plan.
- **[Warmly](https://www.warmly.ai/):** Person-level de-anonymization combined with AI agents for chat, routing, and retargeting across email and ads.
- **[Leadfeeder (Dealfront)](https://www.leadfeeder.com/web-visitors/):** Company-level identification by IP, built around GDPR-conscious defaults, with CRM sync and account alerts.
- **[Clearbit by HubSpot](https://www.hubspot.com/products/clearbit):** Company-level IP intelligence inside HubSpot, showing which companies visit and whether they're already in your CRM.

For pricing, see each vendor's pricing page. We compare them in depth in [RB2B alternatives](https://brandengineering.ai/blog/rb2b-alternatives).

## FAQ

### How can I identify website visitors?

Install a visitor identification script. Account-level tools match IP addresses to companies; person-level tools match visits against an identity graph to return individual contacts. Then route the matches into your CRM with scoring and filtering so sales only sees ICP-fit visitors.

### What is person-level visitor identification?

It identifies the individual behind an anonymous visit, typically returning name, work email, title, and company. Vendors do this by matching cookies, device IDs, and IP addresses against licensed identity data. Coverage is often limited to the US for privacy reasons.

### Is website visitor identification legal?

Generally yes, when done with a lawful basis, proper disclosure, and honoring opt-outs, but the rules differ by region and method. Under GDPR, IP addresses can be personal data, and person-level identification of EU residents gets much more scrutiny. Consult counsel before deploying.

### How accurate is reverse IP lookup?

It works well for visitors on corporate networks and poorly for remote workers, mobile traffic, and VPNs. HubSpot describes its own IP matching as probabilistic. Expect some misattribution and filter heavily by ICP.

### What's the difference between visitor identification and intent data?

Visitor identification tells you who is on your site. Third-party intent data tells you which accounts are researching topics elsewhere on the web. Many teams combine both in account scoring.

## Related

- [RB2B alternatives](https://brandengineering.ai/blog/rb2b-alternatives)
- [What is a GTM engineer?](https://brandengineering.ai/blog/what-is-a-gtm-engineer)
- [GTM Engineer jobs](https://brandengineering.ai/jobs/gtm-engineer)
- [GTM engineer jobs guide](https://brandengineering.ai/blog/gtm-engineer-jobs)
- [Brand Engineering toolkit](https://brandengineering.ai/tools)
