Website Visitor Identification: How to Choose a Tool in 2026

💡TL;DR

Website visitor identification resolves anonymous traffic to a company reliably and to a named person only in narrow, region-dependent cases. Vendors advertise match rates between 20% and 70%, but those numbers are measured on different denominators, so they are not comparable. Choose on four things instead: what the tool resolves, how it proves a match, which regions it legally covers, and whether it writes to your CRM in a shape your reps will act on.

Most B2B sites convert between one and three percent of visitors. Visitor identification software sells the other 97%: the promise that you can see which companies read your pricing page and follow up before they ever fill in a form. The category is real and the good tools work. The buying process, though, is unusually easy to get wrong, because the single number every vendor leads with — match rate — is the least comparable metric in the market.

This guide is about selection, not implementation. It covers what each identification method can actually resolve, how to interrogate a match-rate claim, which vendor categories exist, and the contractual details that decide whether the tool is still in your stack next year. If you have already chosen and want the rollout sequence, read how to identify anonymous website visitors. If your question is about the tracking layer itself — what to collect, retention, cost — see anonymous website visitor tracking.

What visitor identification can and cannot resolve

The category name hides a critical distinction. “Identification” covers two very different outputs with different reliability, legal exposure, and sales value. Deciding which one you are buying is the first and most consequential choice.

Company-level identification. The tool maps a visitor's IP address or network fingerprint to an organisation: “someone at Siemens viewed pricing twice this week”. This is mature, works globally, and carries modest privacy exposure because no individual is named. It is also the only tier most European traffic will ever reach.
Person-level identification. The tool returns an individual: name, work email, sometimes a profile URL. This depends on identity-graph partnerships and consent frameworks that in practice mean US traffic, on a subset of visitors, with meaningful error rates. Coverage outside the US is thin and shrinking.

A third output sits between them and is frequently mis-sold as identification: account-level intent without resolution. Here the tool tells you a company in your target list showed activity across a publisher network, but not that the visit happened on your site. That is third-party intent data, and it belongs in a different budget line. Our guide to intent data providers covers where it fits.

⚠️Reverse-IP has a structural ceiling

Company-level identification depends on a visitor using a corporate network. Remote work, mobile connections, residential ISPs, and VPNs all break that link, and they now account for the majority of B2B browsing. A vendor claiming high company match rates on predominantly mobile traffic is either using a different denominator or inferring from signals it cannot prove. Ask which.

The four identification methods, ranked by how much you can trust them

Every tool in this category combines the same four underlying methods. Knowing which one produced a given match tells you how much weight it deserves, and vendors will tell you if you ask directly.

First-party declared identity. The visitor logged in, clicked an email link with an identifier, or submitted a form. Accuracy approaches certainty and consent is unambiguous. This tier is small but it is the only one worth routing straight to a rep without review.
Reverse IP to organisation. Solid for on-network visitors from medium-to-large companies. Weak for small businesses on consumer ISPs, and it will cheerfully return the name of a coworking space or the ISP itself if the mapping is thin.
Identity graph resolution. A cookie or device is matched against a third-party graph built from publisher and app partnerships. This is what makes person-level claims possible. It is also where accuracy varies most by region and where the compliance questions concentrate.
Probabilistic inference. The tool combines firmographics, behaviour, and timing to guess an account. Useful for prioritisation, dangerous for outreach. Treat any probabilistic match as a research task, never as a verified contact.

A good vendor exposes which method produced each record and gives you a confidence value you can filter on. A weak one returns a flat list with no provenance. That single difference predicts more about whether your sales team will trust the data in month three than any feature on the pricing page.

How to interrogate a match-rate claim

Match rate is a fraction, and vendors choose their own denominator. Before comparing two numbers, force both onto the same basis with four questions.

Percentage of what? All sessions, unique visitors, or only visitors who passed a consent gate? Excluding non-consented traffic can double the headline number without identifying one extra person.
Company or person? A 60% company match and a 60% person match are different products. Vendors sometimes quote the former while demonstrating the latter.
Which regions? Ask for the rate split by US, EU, UK, and rest of world. If a vendor will not split it, assume the aggregate is carried by US traffic.
Verified how? Ask what proportion of matches they would stand behind as confirmed rather than inferred, and how you can tell the two apart in the export.

Then run the only test that matters: a two-week pilot on your own traffic, with your own consent configuration, scored by your own team. Take 50 identified accounts, have a rep check each against what they know, and count how many were both correct and useful. Useful is the harder bar — a correctly identified competitor, recruiter, or existing customer is a true match and a worthless lead.

Comparing the vendor categories

Tools cluster into four categories with different centres of gravity. Shortlist the category first, then compare two vendors inside it; comparing across categories is how buyers end up with a person-level tool for a market where it cannot legally operate.

Platforms in this category — Leadfeeder and Dealfront are the best known, with Albacross and Salespanel competing on price — map visits to companies and surface page-level behaviour per account. Because they never name an individual, they work across Europe and pair naturally with account-based marketing.

The trade-off is that you still have to find the right person inside the identified company, which is a research step the tool does not perform. Budget for that step or the insight stays unused.

Resolves to: Organisation, not individual

Typical pricing: Tiered by identified companies per month

Strongest when: You sell to mid-market and enterprise accounts

Tools such as RB2B built the category by resolving a slice of US traffic to named people in near real time, usually delivered into Slack for immediate follow-up. When the motion fits — US market, short sales cycle, a rep able to act within the hour — the speed advantage is genuine.

Coverage collapses outside the US, and the compliance surface is materially larger than company-level tools. Confirm in writing which jurisdictions are in scope and how opt-outs propagate before you route anything to outreach.

Resolves to: Named individual with work email

Typical pricing: Per identified person, or capped monthly volume

Strongest when: Your buyers are US-based and self-serve

If you already run a marketing automation platform, its de-anonymisation via tracked email links and form fills identifies returning known contacts at effectively zero marginal cost. For companies with a substantial existing database, this often covers more genuinely actionable visits than a new identification vendor would add.

It does nothing for genuinely new traffic, so it is a complement rather than a replacement. Check this option before buying, because many teams pay for identification they already own.

Resolves to: Existing records in your own database

Typical pricing: Bundled into an existing subscription

Strongest when: You have a large email list and long cycles

Broader intent platforms bundle site identification with third-party topic surges across publisher networks. The appeal is one dashboard covering on-site and off-site signals; the risk is that the two get blended, and a rep cannot tell whether an account visited your pricing page or read an unrelated article elsewhere.

If you buy here, insist that on-site and off-site signals stay separately labelled all the way into the CRM. Once merged, the data loses the property that made it credible.

Resolves to: Account-level activity, on and off your site

Typical pricing: Annual platform contracts

Strongest when: You run coordinated ABM across many accounts

✦

Company-level identification tells you which account is interested. Lessie tells you who to email inside it — searching 100+ live sources for the decision-maker who owns the problem, with verified contact details.

Find the right contact free →

Comparison: what each category is worth to you

Category	Resolves to	Region coverage	Privacy exposure	Best fit
Company-level platform	Organisation	Global	Moderate	ABM, EU traffic, mid-market and up
Person-level tool	Named individual	Mostly US	High	Fast US self-serve motions
Automation / CDP add-on	Known contacts only	Global	Low	Large existing database
Intent platform module	Account activity	Varies by vendor	Moderate	Coordinated ABM at scale

Five questions to settle before you sign

These are the terms that decide whether the tool survives its first renewal. None of them appear on a feature comparison, and all of them are negotiable before signature and not after.

Who owns the identified data if you leave? Confirm you can export identified accounts and their activity history, and that the export is self-serve.
How are opt-outs and deletion requests handled? You need a documented path that reaches the vendor's upstream sources, not just their own database.
What happens when volume spikes? Identification is usually metered. Understand the overage behaviour before a campaign produces a surprise invoice.
How does the data land in your CRM? An unstructured note field is useless. Insist on discrete fields for account, confidence, method, page, and timestamp.
What is the contract length and the exit? Prefer a quarterly or monthly term for the first cycle. Annual-only contracts in this category regularly outlive the team's appetite for the workflow.

Where Lessie fits after identification

Identification produces an account, and accounts do not answer email. The step that turns a company name into pipeline is finding the person who owns the problem and reaching them with something specific — and that is a research task no identification vendor performs.

Account to person. Given a company, Lessie searches 100+ live sources to find the roles that matter for your offer, not just whoever appears first in a database.
Verified contact details. Emails are validated before export, so first-touch outreach on a fresh signal does not bounce.
Context for the first line. Recent hiring, product, and public activity give the rep a reason to write that is not “I saw you visited our site” — a line that reliably damages trust.
Works with any identification vendor. Lessie sits after the signal, so it does not lock you into a category choice. See B2B lead finder and buying signals.

The workable shape for most teams is therefore modest: company-level identification for coverage, your existing automation platform for known contacts, and a research layer to make either one actionable. Person-level tooling is worth adding when your market is US-heavy and a rep can genuinely act within the hour — and not before.

FAQ

What is website visitor identification?

Website visitor identification is software that resolves anonymous site traffic to a known entity — either the organisation a visitor belongs to, or in narrower cases a named individual. Company-level identification typically maps a network or IP address to an employer and works globally. Person-level identification relies on third-party identity graphs and consent frameworks, so in practice it covers a subset of United States traffic with materially higher privacy exposure.

How much does visitor identification software cost?

Company-level platforms usually price by the number of identified companies per month, with entry tiers in the low hundreds of dollars and mid-market plans running into four figures monthly. Person-level tools price per identified person or by a capped monthly volume. Because identification is metered, the number that matters is the overage rate: a successful campaign can double traffic and produce an invoice nobody forecast.

Which visitor identification tool is best for European traffic?

Choose a company-level platform. Person-level identification depends on identity graphs whose coverage and legal basis are weak outside the United States, so for EU and UK traffic it will resolve very few visitors and carries the highest compliance exposure. Company-level tools such as Leadfeeder, Dealfront, or Albacross name the organisation rather than an individual, which is both more reliable across Europe and a much smaller privacy surface.

Is website visitor identification GDPR-compliant?

It depends on the tier and the configuration, not on the vendor logo. Company-level identification that names only an organisation is generally the lower-risk option, though the tracking that feeds it still needs a lawful basis and a working consent gate. Person-level identification processes personal data and requires a defensible legal basis, a documented opt-out route that reaches upstream sources, and a data processing agreement. Get your own counsel to review the specific configuration.

Can I trust a vendor’s advertised match rate?

Not without normalising it first. Vendors choose their own denominator, so a rate may be measured against consented visitors only, against sessions rather than unique visitors, or against company matches while the demo showed person matches. Ask for the rate split by region and by resolution tier, then verify with a two-week pilot on your own traffic where a rep scores 50 identified accounts for both accuracy and usefulness.

How do I connect visitor identification to my CRM properly?

Write the signal into discrete fields — account, confidence score, resolution method, page viewed, and timestamp — rather than a free-text note. Reps need to filter on confidence and method to know which records are safe to act on, and merging on-site with third-party intent signals destroys that. Also agree the routing rule before launch, so a single pricing-page visit does not trigger the same follow-up as three visits from two people in one week.