Website visitor identification

How website visitor identification works

Website visitor identification starts with a script on the site. HubSpot's version uses its tracking code to detect the IP address for each page view, then surfaces public information about the company associated with that address 1. Other tools follow the same pattern with their own scripts.

The lookup is the core of the method. HubSpot maintains a proprietary database of company IP addresses 2. Every IP tracking tool keeps its own list of companies and addresses, which is why two tools can show different results for the same traffic 2. The quality of that database decides the match rate a team sees.

The output is a list of companies with the pages they viewed and when. Teams then filter the list by fit, using firmographics, and pass promising accounts to sales or into lead scoring.

Several limits are built into the method.

• Small companies and individuals may not have their own IP address. HubSpot shows these visits as the internet service provider and filters them out by default 1.
• Mobile traffic is hard to place. HubSpot says a visit from a mobile network cannot be tied to a company, while a mobile visit on a corporate network can 2.
• Location can be wrong. HubSpot lists VPNs, provider routing and browsing security software as reasons a visit shows up in the wrong place 2.
• Remote work moves traffic away from office networks, so many visits arrive from home connections that resolve to a provider.

Company-level matching names an organization. Person-level identification goes further and names the individual visitor, usually by tying a device or cookie to identity data. That step depends on identity resolution and carries the most legal risk.

Privacy law covers these identifiers. Recital 30 of the GDPR says people may be associated with online identifiers such as IP addresses and cookie identifiers 3. It adds that these traces, combined with other information, may be used to create profiles of people and identify them 3.

What decides the quality of visitor identification

Four variables shape how useful the output is.

Traffic mix comes first. A site visited mostly from large company offices will match well. A site whose buyers work from home, on mobile or at small firms will see many visits resolve to a provider or to nothing. HubSpot filters provider traffic out by default for this reason 1.

The database comes second. Each vendor's list of company networks is different, so accuracy varies by vendor and by region 2. Test a vendor against accounts known to have visited, such as current customers, before trusting its matches. The same habit applies to any claim of data accuracy.

Page context comes third. A visit to the careers page means something different from a visit to the pricing page. Weight pages by how close they sit to a buying decision and ignore support and job pages.

Consent and law come fourth. Company-level matching and person-level identification carry different risks. Where the GDPR applies, IP addresses and cookie identifiers can count as identifiers that help profile people 3. Person-level tools therefore need a lawful basis and clear notice. Teams should check the rules in each market they sell into with their own counsel.

Example of a visit log and what it can support

A company selling payroll software installs a visitor identification script. After a week, the sales team opens the report.

Company                     Pages viewed                     Visits   Last visit
Example freight company     pricing, payroll integrations    4        Tuesday
Example design studio       blog post                        1        Monday
Internet service provider   pricing, demo                    6        Wednesday   hidden by default
Unknown (mobile network)    pricing                          3        Thursday    no company match

The freight company is the only clear lead. It fits the target profile and viewed pricing and integration pages more than once. A rep can contact it, but the message should not say the rep saw the visit. Mentioning the payroll problem the integration page covers is safer and less unsettling.

The design studio read one blog post. That looks like early research, so it goes into nurture.

The provider row hides real buyers. HubSpot shows visits from small companies and individuals without their own IP addresses as the internet service provider 1. Six visits to pricing and demo pages may come from one interested person at a small firm, but the tool cannot say who.

The mobile row is a dead end. HubSpot says a company cannot be identified when a visit comes from a mobile network 2. In this report, the strongest pricing interest of the week sits in the two rows the tool cannot name.

Teams that want account lists without tracking site visitors take a different route. RichAPI does not offer IP-to-company de-anonymization of website traffic or any intent feed 4. Its signal playbook builds lists from public activity instead, such as job posts, running ads and company posts 4. In that recipe, a search of 200 job posts at 0.1 credits per result bills 20 credits and dedupes to about 60 companies 4. Visitor data shows interest in one seller, while public activity shows what a company did in the open.

Website visitor identification vs public activity signals

Both help find accounts that may be ready to talk. Visitor identification watches traffic on a seller's own site, while public activity signals record what companies did in the open.