De-anonymizing website visitors means matching anonymous traffic to a known organization, almost always by looking up the visitor's IP address, then enriching that company with the firmographics and contacts a rep can actually use. Company-level identification is routine in B2B. Person-level identification is a separate class of product with harder privacy questions, and it is not something Databar does.

This guide covers how identification works and where it fails, which tools do it and what they cost as of September 2026, what Databar can and cannot do here, and what GDPR and CCPA mean for person-level identification.

What does it mean to de-anonymize website visitors?

Your analytics show sessions, pages and traffic sources, not which companies they belong to or who you would contact there. De-anonymizing closes the first gap only. It is three jobs wearing one name:

• Capture. A tag, a tracking pixel or your own server logs record the visitor's IP and the pages they viewed.

• Resolve. The IP is matched against a database of address ranges associated with organizations, returning a company name and domain.

• Enrich and qualify. The company gets firmographics, technographics, contacts and a score, so only the accounts worth a rep's time reach one.

Most teams buy the first two as one product, then find the third is where the work is. "Acme Corp visited" is not a lead. It is a row that still needs a fit score and named people attached.

How does website visitor identification work?

Identification tools split into two groups, and it matters which one you buy.

Company-level tools match the visitor's IP to an organization and return a company name, domain and basic firmographics. That works for people on an office network or corporate VPN, and fails for people at home on a residential connection or on a phone.

Person-level tools match the visit to an individual using identity graphs and cookie data, returning a name, a LinkedIn profile and sometimes an email. Coverage is narrower and the legal footing is different, which is why those vendors restrict it by geography.

Match rates vary a lot by vendor, traffic mix and geography, and published figures are best-case. Run a two-week trial on your own site and count how many identified companies are real and relevant before committing.

Which tools identify website visitors, and what do they cost?

Prices and product facts below were checked on each vendor's own pages September 2026.

One product is no longer on the list: Clearbit Reveal. Clearbit's site now leads with "Clearbit has joined HubSpot!" (checked in September 2026), so if you are not a HubSpot customer, treat Reveal as unavailable. Weighing this layer against broader signal vendors, our comparison of intent data providers covers the ones that bundle third-party intent with first-party visits.

What Databar does and does not do here

Be clear about the boundary, because a lot of content on this topic is not. Databar is not a visitor-identification pixel. There is no tag, nothing to install on your site, no capture of visitor IPs on your behalf. You bring the IPs, from an analytics export, your server logs, or the identification tool you already run.

Databar resolves an IP to the company, not to the individual person behind it.

What is in the catalog is the resolution and enrichment side:

Two caveats on the People Data Labs lookup. It exposes an identified-employee field, but that is best-effort and often empty, so do not assume you will learn who visited. And at 15 credits per address it is for a filtered subset, not every row in a traffic log: the trial's 100 credits buy a handful, enough to judge match quality. Provider details are on the People Data Labs page.

Using Leadmagic for IP-to-company matching

Leadmagic is one of the strongest providers for IP-to-company matching, and it is available through Databar. It gives you the company identification layer that sits in front of everything else in this post.

The setup is a lightweight JavaScript snippet you add to your website. Leadmagic processes visitor IPs against its company database and returns matched organizations with basic firmographic data. From there the pipeline looks like this:

1. A visitor hits your site and the Leadmagic snippet captures the IP and the pages they viewed.

2. The IP is matched to a company, with domain and basic firmographics.

3. Push the matched companies into a Databar table, on a schedule or through a webhook.

4. Enrich each account with the firmographics, technographics and funding data you score on, then find the right people to contact.

5. Route the accounts that clear your bar to the rep who owns them, with the page history attached.

Leadmagic handles the matching. Databar covers the resolution and enrichment side, alongside 160+ other data providers, so the account arrives in your CRM already qualified rather than as a bare company name.

How do you qualify identified visitors against your ICP?

Before spending anything on contacts, score each identified company. Three buckets are enough to start:

Identification tools return thin firmographics. If your ICP depends on a field they do not provide, add it here: a company-by-domain enrichment for size and industry (Explee at 2 credits, Limadata and CompanyEnrich at 3, Surfe at 7.5), a tech stack lookup, a revenue estimate if you segment by it. Technographic data is usually what separates a real fit score from an employee-count filter. On what to score at all, start with lead scoring versus account scoring, and scoring local businesses from their reviews is the same pattern on a different input.

In Databar this filter is a run condition on the enrichment column, written in plain English and translated for you ("only run where country is US and employee count is 50 or more"). Skipped rows do not consume credits, which is the point of putting the filter before the expensive step.

How do you find the right contacts at a company that visited?

For strong-fit accounts, find two or three people in the roles you sell to. With company-level identification you do not know who visited, so aim at the buying group rather than one guess.

Search employees by domain and title (Leadmagic's employee search runs at 0.2 credits, its title search at 6, Explee at 2, Findymail at 3.5), then attach a verified work email to each name. One provider tends to disappoint here, which is why the email column should be a waterfall: providers run in the order you set, the cascade stops at the first usable result, and only the provider that returned data is billed. A lookup that finds nothing is not charged. A partial record is charged, because it did return data.

The setup is covered in how to set up an email waterfall, the choice between cascading and one big database in waterfall versus proprietary enrichment, and the order to put providers in by ContactOut vs PDL vs Diffbot.

Add a verification column after it (Emailable, Bouncer or ZeroBounce, 1 credit each); how email verification works explains the verdicts. Adjacent recipes: contact info from a name and company when a form fill leaves you a name, a reverse email lookup when it leaves only an address, and the phone enrichment roundup for direct dials.

What do the pages they viewed tell you?

Page history is the best free signal you have about buying stage, and most teams throw it away by treating every identified account the same:

A visit is a first-party signal, and it gets stronger next to the ones you already collect. Buying signals and intent data covers combining a pricing-page visit with hiring, funding and technology changes instead of scoring each in isolation.

Is it legal to de-anonymize website visitors under GDPR and CCPA?

General information, not legal advice. The honest answer depends on which of the two products you bought.

Company-level identification is the lower-risk end: you resolve an address to an organization, and organizations are not data subjects. Even so, EU regulators treat IP addresses as personal data in most circumstances, so the processing still needs a lawful basis, a line in your privacy notice, and a retention period you enforce.

Person-level identification is the part people underestimate. Returning a named individual and their contact details from an anonymous visit is processing personal data about someone who never interacted with you, and in the EU and UK the legitimate-interest argument is much harder to make. That is why person-level vendors fence their product geographically. If one offers you person-level identification on EU traffic, ask in writing which lawful basis they rely on and where the consent came from.

Under CCPA and CPRA, IP addresses count as personal information, and buying identity data about California residents can carry disclosure and opt-out obligations. Either way: say what you do in your privacy notice, record which vendor supplied what, exclude consumer traffic, set a retention limit, honour deletion requests. Our guide to privacy-first data enrichment covers the main regimes.

One thing Databar does not do: it offers no consent management, no visitor-side disclosure and no compliance certification for this workflow. Lawful basis, notices and any data processing agreement sit with you and your identification vendor.

How to build the visitor enrichment workflow in Databar

1. Run your identification tool for a week or two before building rules, so you design against real traffic.

2. Write the exclusion list first: competitors, existing customers, partners, your own team, agencies, recruiters, known cloud and ISP ranges.

3. Get the rows into a table. Most identification tools can POST new visitors to a webhook, and a Databar webhook creates one row per request, so nothing needs a manual export. Starting from a file instead, enriching a CSV of companies without code is the same job.

4. If you resolve IPs yourself rather than buying a tool, run the cheap location lookup first to drop datacenter and bot ranges, then the People Data Labs company lookup on what survives.

5. Add firmographic and technographic columns, then put a run condition on the contact columns so only ICP-fit rows spend credits on people.

6. Run the email waterfall and verification on the rows that passed. An AI Researcher column can add a line of company context for the rep to open with.

7. Push the result to HubSpot or Salesforce tagged with the source, pages viewed and visit date. After a month, check which segments produced meetings and tighten the fit rules.

To chain those steps, Flows lays them out on one canvas. To drive it from an agent, the Databar MCP server exposes the same catalog and tables to Claude and other MCP clients, the setup in automating GTM research with Claude Code. For what else fits this stack, see the B2B data enrichment tools roundup.

Mistakes that kill visitor identification programs

Routing every identified visitor to a rep

Reps get a flood of alerts, most of them junk, and within a few weeks they stop opening them. Only strong-fit accounts with high-intent page views should interrupt a rep.

Assuming the person you email is the person who visited

With company-level identification you know the organization, not the individual. The VP you email may never have seen your site. Write outreach relevant to the company's situation, and do not reference the visit as if you watched it.

Enriching before filtering

Running contact enrichment on every identified company, competitors and existing customers included, is the fastest way to burn a budget on records you will never use. The filter belongs before the expensive column.

Questions people ask

Can you identify individual visitors, or only companies?

IP-based identification returns the organization, not the person. Person-level tools do return named individuals, but they rely on identity graphs and cookie data rather than the IP alone, and restrict coverage by geography.

Does Databar identify website visitors?

Not on its own. Databar has no tracking pixel and does not sit on your site. It resolves IP addresses you supply through People Data Labs at 15 credits per lookup, then handles the enrichment, scoring and contact discovery that follows. Capture is your analytics, your server logs, or a dedicated identification tool.

Which IP enrichment maps anonymous traffic to named accounts?

In Databar, "Identify company by IP address" from People Data Labs, at 15 credits. Run the ipinfo.io location lookup first at 0.4 credits to strip datacenter and cloud ranges, so the higher price only hits addresses that could plausibly belong to a company.

What match rate should I expect?

Do not take a number from an article, including this one. Match rates depend on how much of your traffic is office-based, how much is mobile, and which regions it comes from. Trial one tool on two weeks of your own traffic and count the matches that are companies you would sell to.

Do I need consent to identify website visitors?

It depends on your jurisdiction, what the tool collects, and whether it identifies people or only organizations. Company-level sits on firmer ground. Person-level identification on EU or UK traffic needs an answer from your counsel, not a vendor's reassurance.

To try the enrichment side on a week of your own visitor data, start free with the 14-day full-product trial and 100 credits, or book a founder demo and we will build the table with you. Paid plans start at $99/month with 5,000 credits, and lookups that return nothing are not charged.