Company-Level vs Person-Level Visitor Identification

Visitor identification tools promise to tell a B2B team who is on its website. They answer that question at two very different depths, and the depth you choose shapes coverage, accuracy, legal exposure and what sales does next.

This applies to B2B sites with enough traffic to identify. The person-level case also needs a meaningful share of US visitors and a sales team that acts within days. Every published match rate is vendor-measured on its own denominator, so a trial on your own traffic is the only reliable number.

Company-Level Vs Person-Level Visitor Identification

Company-level visitor identification names the business behind a website visit, while person-level identification names the individual, and vendors such as RB2B and Vector limit person-level matching to US visitors. That one difference drives every other trade-off in this comparison.

For most B2B teams, company-level identification is the default. It is not limited to US visitors (Snitcher describes its company-level identification as global, and RB2B runs global company-level matching through Demandbase), it fits deals that a buying committee decides, and it rests on firmer ground under European privacy rules. Person-level identification adds named US visitors on top, at a higher privacy and accuracy risk.

Read the table as a pair of trade-offs, not a winner. Company-level coverage is wider, but the record stops at the account. Person-level gives sales a name, but only for US visitors, only as a probability, and only where the pixel can load lawfully.

The legal row matters most for teams with European traffic. The European Data Protection Board treats tracking pixels as access that needs consent, while access to an IP address triggers the same rule only in certain cases. That gap is why company-level tools built on IP lookup can serve EU traffic more easily than person-level pixels.

The workflow row matters for everyone. Forrester's buyer research describes typical B2B decisions as involving a large group of internal stakeholders and outside influencers. Even a correct person-level match usually names one member of that group, not the buyer.

What Is The Difference Between Company-Level And Person-Level Visitor Identification?

The difference is the record each method returns, since Snitcher states it identifies the companies visiting a site and not individual consumers, while RB2B returns a visitor's name, LinkedIn profile and business email. One answers which account is interested. The other answers which person is interested, and only for US visitors.

Company-level identification is the established form of website visitor identification. It maps a visit to the business behind it. Leadfeeder (formerly Dealfront) describes its output as which companies are on a site, how well they fit the ideal customer profile and how interested they are. The category guide explains how company matching works in more depth.

Person-level identification goes one step further and names the individual. RB2B returns names, LinkedIn profiles and business emails, and Vector returns a name, role and work email. Both restrict that match to US visitors.

Company-level tools often show people too, which causes confusion. Leadfeeder enriches identified companies with contacts, and Snitcher enriches accounts with decision-maker contacts. Those contacts work at the visiting company. They are not the person who visited.

Treating an enrichment contact as the visitor leads to outreach that mentions a visit the recipient never made. Treat the company as the signal and the contacts as a list of people who might care about it.

How Does Company-Level Vs Person-Level Identification Work?

Person-level tools match a visitor's device against an identity graph, and RB2B says its database holds only profiles matched to US home addresses of consumers who agreed to third-party marketing. Company-level tools work differently. They look up the visitor's IP address in a database that maps network addresses to the companies that use them.

Company-level, by IP intelligence: Leadfeeder says its IP database receives a constant stream of updates from multiple data sources. Its own blog reports more than 45 million IP-to-company mappings covering about 9 million companies. Snitcher processes the visitor's IP address to look up the company and does not share that address with its customers.

The weak point is the home office. Leadfeeder's own article says company-level identification struggles with remote workers on home broadband, because a home connection does not map to an employer.

Person-level, by identity graph: Vector says its matching compares a device fingerprint against an identity graph. RB2B's compliance page explains where the names come from. Its database only contains profiles matched to US home addresses, it uses IP ringfencing to resolve only US traffic, and the consumers in it have agreed to provide their information for third-party marketing. Retention, the identity-data company behind RB2B, hosts a consumer opt-out page that customers can link to.

That sourcing matters. A person-level name comes from a consumer marketing database matched to a home address, not from anything the visitor typed on your site. Leadfeeder's article also says person-level matching fails with VPNs, Safari tracking protection and Apple Private Relay.

A probability, not a fact: Retention's APIs return hashed emails with confidence scores and up to three ranked matches per visitor. Vector states that identification is probabilistic and that coverage will vary. A person-level match is the most likely candidate for a device, so sales should treat it as a lead to check, not a confirmed identity.

Known-person identification, a third route: Snitcher names an individual only when that person provides an email, through an auto-detected form submission, an identify call at login or signup, or a click on an email link that carries the address. Earlier anonymous sessions are then linked to that person while cookies persist. Vector also accepts known contacts through an identify method or form submissions.

Known-person identification is not deanonymisation. The name comes from an email the visitor supplied, not from a database match.

What Match Rates Can You Expect From Company-Level And Person-Level Identification?

Vector's own help centre puts its contact-level match rate at about 10% of traffic and about 30% with company-level fallback, with 60 to 75% accuracy against known HubSpot contacts, so a person-level match is a probability rather than a fact. Published company-level rates use different denominators and regions, so they do not compare directly with Vector's.

Company-level, vendor-stated: Leadfeeder reports from its own data that it identified 38.0% of visits across five core regions in August 2026, and 30.8% of visits in the US and Canada.

The denominator problem: Knock2, a vendor not covered here, reports that 93% of engaged sessions resolved to a company and 62% resolved to a person on US engaged sessions. Knock2 counts a session as engaged after 10 or more seconds or two or more pageviews. Removing short visits from the denominator raises the reported rate, so Knock2's 93% and Leadfeeder's 38.0% measure different things.

Published ranges conflict: Happierleads' comparison article gives person-level match rates of 5 to 20% in its body and 30 to 55% in its FAQ, with no source for either range.

Vendor pages disagree with themselves: Leadfeeder's homepage says it can identify companies even when their employees work remotely. Its own blog article says company-level identification struggles with remote workers on home broadband. A homepage claim is no substitute for a test.

Match rate and accuracy are separate numbers. A match rate counts how many visits returned a record. Accuracy counts how many of those records were right. Vector publishes both, and its 60 to 75% accuracy band means a quarter or more of its contact matches did not agree with known HubSpot contacts in its own tests.

Before comparing vendors on match rate, ask each one these questions:

• Is the figure a share of all visits or of engaged sessions only?
• Which regions does the figure cover?
• Is it a match rate or an accuracy rate?
• Can we run a trial on our own traffic before signing?

Is Person-Level Visitor Identification Legal?

Person-level identification is legal in narrower conditions than company-level, because the European Data Protection Board's guidelines on Article 5(3) of the ePrivacy Directive treat tracking pixels as access that needs consent, and US vendors such as RB2B resolve only US traffic. In the US, state law rather than GDPR sets the limits.

In the EU: The EDPB's Guidelines 2/2023, final version of 16 October 2024, say tracking pixels and tracking links fall under Article 5(3) "even if this storage is not permanent". They need consent unless an exemption applies. The same guidelines say access to IP addresses triggers Article 5(3) only in certain cases, which is the footing company-level tools built on IP lookup argue from.

Enforcement under GDPR is substantial. A DLA Piper survey of 16 European jurisdictions found that aggregate GDPR fines from May 2018 to January 2026 total EUR 7.1 billion, with about EUR 1.2 billion issued in 2025.

In the US: RB2B's compliance page says GDPR does not apply to its person-level product, because it uses IP ringfencing to resolve only US traffic. That is the vendor's legal position, not a regulator's. RB2B also says state laws such as CCPA and CPRA may apply to customers that handle substantial data.

California's Governor signed SB 690 on 30 September 2026. According to Mayer Brown's analysis, SB 690 removes private lawsuits under CIPA's pen register and trap-and-trace provisions as applied to websites and apps from 1 January 2027. Those claims had carried statutory damages of at least USD 5,000 per violation.

Wiretap claims remain. Private suits under CIPA's wiretapping and eavesdropping provisions are still possible where the data is the contents of a communication, such as a submitted web form.

Vendor positions differ by design:

• Snitcher: Relies on legitimate interest under Article 6(1)(f) GDPR, processes and stores data in Frankfurt, and offers a consent mode that sets no cookies or persistent storage until consent. It integrates with Cookiebot, OneTrust and Transcend.
• Leadfeeder: Processes and stores all data within the EU.
• RB2B: Runs global company-level identification through Demandbase and says consent capture is required for GDPR on that traffic.
• Vector: Says its pixel should be categorised in the site's consent-management setup, with the legal or privacy team deciding when it loads, and that identification respects the consent banner.

Whichever approach you choose, plan to install a website visitor identification tool behind a consent banner on pages that EU visitors reach.

This section is general information, not legal advice.

How Do Sales And Marketing Teams Act On Each Signal?

A company-level signal suits how B2B teams actually buy, since Forrester's State of Business Buying report counts 13 internal stakeholders and nine external influencers in a typical buying decision, so one named visitor rarely is the buyer. A person-level signal suits a different motion, a fast one-to-one follow-up with someone who showed interest.

Buyers also prefer to move without sellers. A Gartner survey of 646 B2B buyers, run from August to September 2025, found that 67% of B2B buyers prefer a rep-free experience, and 45% used AI during a recent purchase. Both figures come from an independent survey.

That is an analyst's opinion, separate from the survey numbers. Barbara Winters, vice president and principal analyst at Forrester, made a related point, saying that B2B buyers are under immense pressure to justify investments and minimize risk.

Acting on a company signal: Score the account against the ideal customer profile, alert its owner and work the wider buying group with content and ads. Teams that route identified accounts to sales set ownership and follow-up rules first, so a visit reaches the right rep.

Acting on a person signal: Reach the named visitor quickly, and check the match before referencing it. A wrong match turns personal outreach into an awkward cold message.

Where each tool sends its signal:

• RB2B: Slack, Teams, CRM and Clay.
• Leadfeeder: Salesforce, HubSpot, Pipedrive, Microsoft Dynamics, Zoho and Slack.
• Vector: Salesforce, HubSpot and Pardot, a single-channel Slack bot, and Zapier and Clay through webhooks.

The trade-off is between reach and precision. A company signal reaches the whole committee without naming anyone. A person signal names someone, but only one person, and only a share of the time.

Company-Level And Person-Level Tools Side By Side

Of the five tools compared, Leadfeeder and Snitcher identify companies, RB2B and Vector identify US individuals, and Retention supplies the person-level identity data behind RB2B. The five represent both approaches, two company-level tools, two person-level tools and one identity-data provider.

Each row compares identification level, where person-level matching works, the data returned and where it goes. Rows reflect vendor documentation, plus one independent review for RB2B, and every match rate in this post is vendor-stated.

The table is a map of approaches, not a ranking. Each tool's section below covers the same points in the same order.

Leadfeeder

Leadfeeder is a company-level tool that shows which businesses visit a site and how well they fit the ideal customer profile, and it processes and stores its data within the EU.

Level and method: Company-level, through an IP database the vendor says receives a constant stream of updates from multiple data sources.

Geography: Leadfeeder's own data covers five core regions, where it identified 38.0% of visits in August 2026, and 30.8% of visits in the US and Canada.

What it returns: Which companies visit, how well they fit the ideal customer profile and how interested they are, plus enrichment contacts at those companies.

Where data goes: Salesforce, HubSpot, Pipedrive, Microsoft Dynamics, Zoho and Slack.

Privacy position: All data is processed and stored within the EU.

What drives cost: The plan tier. Leadfeeder says its return from the Dealfront name changed no prices, and its plans are in the pricing table below.

Limitation: Leadfeeder's own article says company-level identification struggles with remote workers on home broadband.

Best fit: Teams that want EU data storage, sell to buying committees and want account signals synced to the CRM. Not a fit for teams that need named individual visitors or whose buyers mostly browse from home broadband.

Snitcher

Snitcher identifies companies rather than individual consumers and names a person only after that person submits an email through a form, a login or a tracked email link.

Level and method: Company-level, by IP intelligence. The visitor's IP address is used to look up the company and is not shared with customers.

Geography: Global identification at company level, by the vendor's own description.

What it returns: Industry, headcount, revenue, tech stack and headquarters for each identified company, with decision-maker contacts as account enrichment.

Where data goes: Slack, CRM records or outreach sequences.

Privacy position: Legitimate interest under Article 6(1)(f) GDPR, data processed and stored in Frankfurt, and a consent mode that sets no cookies or persistent storage until consent.

What drives cost: The number of companies identified, which sets the plan tier shown in the pricing table below.

Limitation: No anonymous person-level matching. A visitor is named only after giving an email.

Best fit: EU-facing teams that want account signals plus named contacts from their own forms, logins and emails. Not a fit for teams that want anonymous visitors named without a form, login or email click.

RB2B

RB2B names individual US visitors with their LinkedIn profiles and business emails and sends them to Slack, Teams or a CRM, while visitors outside the US are matched only to their company.

Level and method: Person-level for US visitors from a US-only database, plus company-level matching through Demandbase.

Geography: Person-level is US only, enforced by IP ringfencing. Company-level is global, and RB2B says that needs consent capture for GDPR.

What it returns: Names, LinkedIn profiles and business emails for US visitors, and the company for visitors elsewhere.

Where data goes: Slack, Teams, CRM and Clay.

Privacy position: RB2B says GDPR does not apply to its US-only person-level product and that state laws such as CCPA and CPRA may apply to customers handling substantial data. Its database holds consumers who agreed to third-party marketing, with an opt-out page hosted by Retention.com.

What drives cost: Resolution-based pricing, so the bill follows how many visitors RB2B identifies each month, according to an independent review by Salesforge. The plans are in the pricing table below.

Limitation: The same review, from a publisher that sells a competing product, reports complaints that some profiles are incomplete or inaccurate and some leads are irrelevant and need filtering. It also notes no historical data before setup and limited value on low-traffic sites.

Best fit: US-heavy B2B sites with an SDR team that follows up with named individuals within days. Not a fit for low-traffic sites or sites whose visitors are mostly outside the US.

Vector

Vector resolves US visitors to a name, role and work email and falls back to company-level matching, and its help centre describes identification as probabilistic with coverage that varies.

Level and method: Contact-level and account-level, matching a device fingerprint against an identity graph, with IP-to-company matching only as a fallback.

Geography: Geofenced to the US, including the company-level fallback.

What it returns: The visitor's name, role, work email and pages read, which Vector calls a smaller, higher-intent slice of B2B traffic.

Where data goes: HubSpot, Salesforce and Pardot, a single-channel Slack bot, and Zapier and Clay through webhooks. SiteID pushes every 30 minutes.

Privacy position: Vector says identification respects the consent banner, each record has a documented source and retention window, and the legal or privacy team should decide when the pixel loads.

What drives cost: A monthly identification cap that pauses identification when reached, with no automatic overages, plus a separate monthly limit on ad audiences, CSV exports included. The plan is in the pricing table below.

Limitation: Vector's own figures put contact matches at about 10% of traffic and accuracy at 60 to 75% against cookied HubSpot contacts.

Best fit: US account-based marketing teams that want named contacts in HubSpot or Salesforce and in ad audiences. Not a fit for teams with mostly non-US traffic, since even its company-level fallback is geofenced to the US.

Retention

Retention resolves anonymous website traffic to a person and returns up to three ranked hashed-email matches with confidence scores, which makes plain that person-level identification is a ranked guess.

Level and method: Person-level identity resolution, which Retention describes as resolving traffic to a person with consent-aware signals.

Geography: US identities, with more than 250 million US identities and more than 600 individual attributes, by Retention's own count.

What it returns: MD5 and SHA256 hashed emails with confidence scores, plus a B2B professional identity (name, title, seniority, department, firmographics, LinkedIn URL and work email) or a B2C personal identity (hashed and personal emails and device IDs).

Where data goes: Real-time APIs or bulk and batch files, with batch lookups of up to 1,000 records.

Privacy position: RB2B's compliance page describes the database as consumers who agreed to provide their information for third-party marketing, and Retention.com hosts the consumer opt-out page.

Limitation: Matches arrive as ranked candidates rather than one confirmed identity, and several capabilities on its API page are announced as in development and not yet available.

Best fit: DTC ecommerce teams growing an email list, and teams that want identity data by API rather than a finished sales tool. Not a fit for sales teams that need one confirmed identity per visitor, since matches arrive as ranked candidates.

What Does Company-Level Vs Person-Level Identification Cost To Run?

RB2B bills by the visitors it identifies each month, according to Salesforge's review, and Vector caps identifications per month on its plan and pauses at the cap, so person-level running cost tracks identified traffic more than seat count. Company-level tools sell plans, which the pricing tables in the Leadfeeder and Snitcher sections show.

Person-level limits: RB2B uses resolution-based pricing, so the bill follows how many visitors it identifies each month, according to Salesforge's review. Vector sells a plan, shown in its pricing table, that counts identified visitors toward a monthly cap and pauses identification at the cap, with no automatic overages. Its ad audiences, CSV exports included, have a separate monthly limit.

Those limits have two consequences. RB2B's bill rises with identified traffic, not with the number of people using the tool. And on Vector, a traffic spike can use up the cap before the month ends, leaving later visitors unidentified until it resets.

Company-level plans: Leadfeeder's return from the Dealfront name changed no prices, according to the vendor. Its pricing table and Snitcher's show the plans to compare.

Costs that sit outside the plan apply to both approaches:

• Consent management for EU-facing pages, so each pixel loads only when it lawfully can.
• Time spent filtering person-level leads, since the independent RB2B review reports irrelevant leads that need filtering.
• Routing work in the CRM and Slack, so either signal reaches someone who acts on it.

Decision rule: Size a person-level plan from your US traffic share before buying. Only US visitors can be named, so a site with mostly EU traffic buys capacity it can barely use.

Can You Use Company-Level And Person-Level Identification Together?

Yes. One setup runs company-level identification on all traffic and adds person-level matching for US visitors, which is how RB2B pairs its US person-level ID with global company-level ID. Vector also falls back to company-level matching for US visitors when no contact match is available.

The same tool can do both, in three common patterns:

• One tool, two levels: RB2B names US visitors and matches everyone else to a company through Demandbase. Vector returns contacts where it can and companies where it cannot, within the US.
• Company-level plus known people: Snitcher stays company-level for anonymous traffic and names a person when that person submits a form, logs in or clicks a tracked email link, linking earlier sessions while cookies persist.
• Two tools side by side: A company-level tool runs on all traffic, and a person-level pixel loads only for US visitors, behind consent.

Switching from company-level to person-level: Three things change. The consent setup must now cover a tracking pixel, which EU guidance treats as needing consent. Routing moves from account owners to fast individual follow-up. Cost starts to track identified visitors, through RB2B's resolution-based pricing or Vector's monthly identification cap.

Switching from person-level back to company-level: Named US visitors disappear from alerts, so sales returns to working accounts and their buying groups. Known-person identification through forms, logins and email clicks, as Snitcher offers, keeps a name attached to the visitors who chose to share one.

Which One Should You Choose?

Choose company-level identification when a meaningful share of traffic comes from the EU or deals involve a buying committee, and add person-level identification only for US traffic when sales follows up with individuals within days. The rest of the decision comes down to where traffic comes from and who acts on the signal.

Picks by situation:

• EU traffic, committee buying: Company-level, with Leadfeeder for EU data storage or Snitcher for global company-level identification. Pick Snitcher when you also want names from your own forms and email clicks.
• US-heavy traffic, short sales cycle, SDR follow-up: RB2B, because it names US visitors and pushes them straight to Slack, Teams or the CRM.
• US account-based marketing with ad audiences: Vector, because it feeds named contacts into HubSpot, Salesforce and ad audiences.
• DTC ecommerce growing an email list: Retention, through Retention FLEX.
• Mixed US and EU traffic: Company-level everywhere, plus person-level for US visitors only, behind consent.

Three shifts favour company-level as the base layer. Gartner's survey found most B2B buyers prefer a rep-free experience, which makes anonymous account signals more useful than a cold call. Forrester counts large buying groups behind each decision. And EU guidance treats pixels as needing consent, while California's SB 690 removes one class of private website lawsuits from 2027 but leaves wiretap claims over form contents in place.

• Test each tool on your own traffic before signing
• Ask every vendor for the denominator behind its match rate
• Check your US traffic share before buying a person-level plan
• Confirm the consent setup for each pixel with counsel
• Decide who acts on account signals and who acts on named visitors

For tool picks within each approach, compare the best company-level visitor identification tools and the best person-level visitor identification tools.

Frequently Asked Questions

Sources