To identify website visitors, B2B teams use two methods: a reverse IP lookup that guesses the visitor's company from their network, or first-party signals where the visitor tells you who they are, by asking a question, booking a meeting or replying to an email. The first gives you company names for part of your traffic; the second gives you fewer names, but with the reason behind each visit.
This guide is part of our series on buyer intent signals in B2B. It explains how each method works, why IP identification misses so many visits, what the privacy rules say in the United States and in Europe, and how to combine both without writing to people who never asked to hear from you.
Why do B2B teams want to identify website visitors?
Most B2B visitors leave without a trace. First Page Sage puts the average conversion rate of a B2B SaaS website at 1.1%, with the highest industry in its table, legal services, at 7.4% (First Page Sage). The other 90-plus percent read your pricing page, compare you with two competitors and close the tab.
Sales teams see those visits in analytics as numbers. They want names: which accounts are researching, which ones came back three times this week, and who to call first. That is the job visitor identification tools promise to do. If you want a rough figure for your own site, you can estimate what your lost visitors cost you.
How do reverse IP tools identify website visitors?
Every visit comes from an IP address. A reverse IP tool matches that address against a database of networks owned by companies. If the address belongs to the office network of a known company, the tool shows that company in a list of visitors, with the pages viewed.
Leadfeeder, one of the best-known tools, describes it plainly: it "matches the associated IP address against a continuously updated company database", and all tracking is at the company level (Leadfeeder). CRMs include the same idea. HubSpot updates tracked visits for target companies "using HubSpot's reverse-IP technology" (HubSpot Knowledge Base).
Company-level vs person-level identification
Company-level tools tell you that someone at a given company visited. Person-level tools go further and claim to name the individual. They do not rely on the IP address alone. RB2B, for example, explains that it matches visitors against a publisher network "using a combination of 1st and 3rd party cookies, device IDs and IP addresses", and states that its person-level identification is US-only (RB2B).
What match rates do vendors claim?
Each vendor publishes its own number. Leadfeeder says it identifies up to 45% of B2B visitors at company level. RB2B says it identifies 70 to 80% of website traffic. These are marketing claims measured on the vendors' own terms, and they vary with your audience. A site that sells to large enterprises with central networks will see far more matches than one that sells to small teams who work remotely.
Why does IP identification miss so many visitors?
An IP address identifies a network, not a person. The match only works when the network clearly belongs to one company. Many B2B visits do not meet that condition:
- Remote and hybrid work. A buyer reading your pricing page from home uses their internet provider's address, which points to the provider, not the employer.
- Mobile data. Phones on a carrier network share addresses with thousands of other users.
- VPNs and privacy relays. With iCloud Private Relay, a second relay "generates a temporary IP address" before the request reaches the website, so the site does not see the user's own address (Apple).
- Shared offices and small companies. Coworking spaces and small firms rarely own a network registered in their name.
There is a second limit, which matters more for sales. Even a correct match tells you the company, not the person, and not the reason. Someone from a 5,000-person company visited your careers page: is that a buyer, a candidate or a competitor? The tool cannot say.
Your analytics will not fill the gap either. Google states that in Google Analytics 4 "IP addresses are not logged or stored" (Google Analytics Help).
Is it legal to identify anonymous website visitors?
The answer depends on where your visitors are.
In Europe, the GDPR names "internet protocol addresses" among the online identifiers that can make a person identifiable (GDPR, Recital 30). Identification tools also place or read trackers. In France, the CNIL requires visitors to be informed and to give consent before most trackers are placed or read; the exceptions cover trackers that are strictly necessary for the service, such as a shopping cart or authentication (CNIL). RB2B itself notes that consent capture is required under the GDPR for its global company-level identification.
In the United States, there is no single federal rule for visitor tracking, but state laws apply. California's CCPA covers information that "could reasonably be linked with you or your household", including internet browsing history (California Attorney General). And if you email someone you identified, CAN-SPAM applies: "The law makes no exception for business-to-business email", and opt-outs must be honored within 10 business days (FTC).
This guide is not legal advice. The practical rule most teams adopt: use company-level data, if at all, to prioritize accounts, and send individual messages only to people who gave you their details.
What are first-party signals?
First-party signals are what buyers do on your own channels, with their knowledge. The useful ones for sales are:
- The questions they ask. A visitor who asks "do you integrate with our scheduling tool?" has told you their stage and their objection.
- The pages they read. Pricing, security, integrations and case studies point to evaluation, not curiosity.
- Return visits. A third visit to the pricing page in a week means more than one long session.
- What they share on purpose. An email left in a conversation, a meeting booked, a reply to a follow-up.
These signals identify fewer visitors than an IP lookup claims to. But each one comes with a name the visitor chose to give and with the reason they came. We list more of them, with examples, in our article on B2B buying signals.
This is where a conversation on the site changes the picture. An agent that answers visitors while they read turns an anonymous session into a person with a question. Nudge's Outreach Agent then follows up with those who come back without writing, for example on their third visit to the pricing page, and prepares the message, so you do not need to buy a list.
How to identify website visitors with first-party signals
You can set this up in six steps, whatever tool you use.
1. Give visitors a reason to talk
Forms ask for details before giving anything. A chat or an AI agent that answers questions first gets the visitor talking, and people share an email more easily once they got a useful answer.
2. Ask for details at the right moment
Ask for an email when there is something to send: a meeting confirmation, a document, an answer that needs a human. Not as a gate before the first reply.
3. Link page views to the conversation
Once a visitor identifies themselves, connect their earlier and later visits to that conversation. You know what they asked and what they read afterwards.
4. Pick a few signals worth acting on
Choose three or four: a return to pricing, a question about integrations or security, a meeting page left halfway. Ignore the rest at first.
5. Follow up from the question, not from the tracking
Write about what the buyer asked, not about the fact that you saw them. "You asked about onboarding time: here is how it works for a team of 20" works. "I noticed you visited our pricing page" does not.
6. Keep IP data for prioritization only
If you also use a company-level tool, use it to rank accounts for your team, and keep it out of individual cold emails. Our guide to warm outbound prospecting covers how to write to buyers who already showed interest.
IP lookup vs first-party signals: which should you use?
Use reverse IP lookup when you sell to large accounts with known networks, run account-based marketing on a fixed target list, and mostly need to know which accounts are active. Expect partial coverage, and check consent rules if you have European traffic.
Use first-party signals when you want to talk to the actual buyer, need to know why they came, and want to stay clear of privacy issues. Coverage depends on how many visitors you get talking, which is why the quality of the AI agent on your website matters more than the tracking script.
Many teams combine them: IP data to see which accounts are warming up, and conversations to find the person and the need.
Example: from an anonymous visit to a booked meeting
A visitor lands on a B2B software site on a Tuesday evening, from a home connection. A reverse IP tool would show their internet provider, not their employer.
They open the pricing page and ask the site's agent: "Is there a minimum number of seats?" The agent answers from the pricing page and cites it. The visitor then asks whether onboarding can be done in French. The agent answers, offers a 20-minute slot with a sales rep, and the visitor books it, leaving their work email.
The sales rep now knows the company, the person, team size, language and the first question. No IP lookup was involved, and nobody received an email they did not expect.
How Nudge handles visitor identification
Nudge builds conversational AI agents for B2B sales teams. The Website Agent answers visitors while they read, trained on your site and documents, and each answer cites the page it comes from. When the visitor is ready, it books the meeting inside the conversation, without a form.
For visitors who leave and come back, the Outreach Agent follows up with those who return without writing, for example on a third visit to the pricing page, and prepares the message, so there is no list to buy. The Nurturing Agent picks up conversations where they stopped: each follow-up starts from the visitor's own question. The Handoff Agent then gives the closer a brief with what the prospect wants, what they asked and the objection already raised.
See how it works on the Outreach Agent page, or ask the agent on our site a question.
Frequently asked questions
Can you see who visited your website?
Not by default. Analytics tools show sessions, pages and sources, but not names. Google Analytics 4 does not log or store IP addresses at all. To put a name on a visit, you either use a visitor identification tool that guesses the company from the IP address, or you give visitors a reason to identify themselves: a question answered in a chat, a meeting booked, a form or an email link.
How accurate is reverse IP lookup?
It depends on your audience. Lookup works well for visitors on a large company's own network and poorly for people working from home, on mobile data, behind a VPN or using a privacy relay. Vendors publish their own figures: Leadfeeder says it identifies up to 45% of B2B visitors at company level. Test any tool on your own traffic before you trust a match rate.
Is identifying website visitors legal?
It depends on where your visitors are and how you identify them. Under the GDPR an IP address can be personal data, and in France the CNIL requires consent before most trackers are read or placed. In the United States, California's CCPA covers information that could reasonably be linked to a person. Person-level identification tools generally limit themselves to US visitors for this reason.
What is the difference between company-level and person-level identification?
Company-level identification tells you that someone from a given company visited, based on the network they used. Person-level identification claims to name the individual, usually by matching cookies, device IDs and IP addresses against a data network. RB2B, for example, states that its person-level identification is US-only. Neither tells you why the person came, which is what a first-party conversation adds.
What are first-party signals?
First-party signals are the actions buyers take on your own channels: the pages they read, how often they come back, the question they ask in a chat, the meeting they book or the email they reply to. You collect them directly, with the visitor's knowledge, so they are more reliable than inferred data and easier to act on without breaking privacy rules.




