Identifying Who Clicks Your LinkedIn Ads

LinkedIn's Insight Tag reports a click. It does not report a person. The distinction matters because everything a B2B marketer wants to do after the click (qualify the lead, route it to sales, personalize the follow-up) depends on knowing who clicked. This article explains how person-level visitor identification picks up where the Insight Tag stops, and what changes once that gap is closed.

The Insight Tag is a JavaScript snippet placed in a site's footer. It collects IP address, time-on-page, and activity data via cookies, then LinkedIn matches that data against its member profiles. The output is aggregate demographic information: job title, seniority, industry, company size, and location. It never resolves to a named individual. Retargeting through the tag reaches any website visitor who has a LinkedIn cookie in their browser, including people who arrived from organic search, direct traffic, or other paid channels. But matching is incomplete, because not every visitor is a recognized LinkedIn member account.

Two structural limits compound each other. Privacy restrictions and third-party cookie loss are why the tag cannot match every visitor, even within its intended scope. Practitioners acknowledge this: DemandSense's retargeting guide recommends pairing the Insight Tag with LinkedIn's Conversions API, a server-to-server connection, specifically to recover conversions the tag misses. The tag alone is understood to be partial.

The practical consequence is concrete. A visitor spends four minutes on a pricing page after clicking a LinkedIn ad, reads case studies, and leaves. The advertiser sees one click. The person is unknown. That gap is not small. Real person-level match rates for B2B website visitor identification run 8–41%, so you should scrutinize most vendor claims before you buy.

Why company-level identification is not enough for LinkedIn ad follow-up

Knowing that Acme Corp visited a pricing page is interesting. Knowing which person at Acme visited is the only version of the data a sales team can act on. Company-level identification is the first and most common fix for the Insight Tag's blind spot, and it is genuinely useful for some purposes. It is insufficient for the follow-up motion LinkedIn ad teams actually want.

Company-level identification works through reverse IP lookup. The tool associates a visitor's IP address with a corporate network and returns firmographic data: industry, company size, revenue. What it does not return is the individual. This approach works globally at the company level, but match rates fall for EU traffic. It is generally defensible under GDPR via legitimate interest, provided a data processing agreement, a privacy-policy disclosure, and an opt-out route are in place. The compliance path exists, which is why company-level tools remain the default for international coverage. Warmly claims 65% company-level identification via 20+ data providers, and that number tells you how far reverse-IP infrastructure has come.

The gap matters most for LinkedIn ad traffic specifically. The click already tells you the ad resonated with someone. What is missing is not the account. It is the individual. Company-level data supports account prioritization and retargeting audience construction. A sales rep, however, cannot personalize an outreach message, choose the right LinkedIn connection, or time a call without a name and title. A company match tells you where to look. It does not tell you who to call.

How person-level identification closes the gap between a click and a named contact

Person-level identification resolves an anonymous browser session to a named individual by cross-referencing multiple signals against an identity graph. It does not extend what the Insight Tag does. It is a different mechanism entirely. The signals involved include IP address, cookie data, device fingerprints, and, where available, hashed email identifiers and advertising-network data. These are stitched together against identity graphs and data enrichment databases. A stable identifier, such as a hashed email, is what makes a match deterministic rather than guessed.

The two matching approaches behave very differently. Deterministic matching ties a session to a known identity when a stable, verified identifier is present. When it fires, confidence is high. Probabilistic matching scores a bundle of weak signals — IP range, device type, browsing pattern, time of day — and returns a best guess when the score crosses a threshold. A score at that level of confidence still means a meaningful share of identifications are wrong, and nothing in the output flags which ones. The distinction matters because vendors rarely volunteer which method produced a given record.

When identification fires, the output carries name, company, job title, LinkedIn profile URL, and work email. Those are the fields that let a sales rep send a relevant, timely message. Richer implementations add seniority, job function, visit count, pages viewed, scroll depth, traffic source, and which ad platform brought the visitor. The geographic constraint is hard and should not be negotiated: person-level identification is permissible for U.S. traffic under CCPA/CPRA, but for EU and UK visitors GDPR effectively limits identification to company-level because person-level matching generally requires consent that almost no visitor grants. Honest vendors enforce this automatically. For enterprise visitors from offices, expected company-level match rates run 50–65%.

What match rates actually measure and why vendor figures are unreliable

Match rate is the most gamed figure in visitor identification. The same term covers three different measurements, and vendors choose between them based on which looks best in a sales conversation. Coverage measures what share of total visitors get identified at all. Resolution level distinguishes whether an identification means a company name or an actual person. Accuracy asks how many of the identifications made are correct and complete. A vendor may quote 65% for company coverage, but you may have assumed it meant person-level accuracy.

Vendors also choose their own denominator. A rate may be measured against consented visitors only, against sessions rather than unique visitors, or against company matches even when the demo showed person matches. The spread is not hypothetical. RB2B has been published at four different person-level match rates within the same year, with figures ranging across a roughly five- to nine-fold spread depending on who measured and what denominator they used. Real-world benchmarks in 2026 run well short of most vendor advertising at both the company and person level. LinkedIn paid traffic reaches higher company-level rates than organic traffic, but mobile-heavy usage patterns constrain it. SMB visitors from offices have 30–45% expected company-level match rates, and international traffic sits at 15–30%.

A reliable self-audit avoids these distortions. Use a fixed 30-day window. Count unique U.S. visitors, not sessions or pageviews, as the denominator. Export and deduplicate identified people. Keep person-level and company-level results on separate lines. Sample identified contacts to check email deliverability and current role. One additional distortion matters more now than it did two years ago: AI crawlers and bots inflate raw visit counts. GPTBot, ClaudeBot, PerplexityBot, and Google-Extended are among the most active, and most website owners cannot see them in standard analytics. A tool that does not separate bot and AI agent sessions from human sessions is reporting a flattering denominator. A match rate calculated against total sessions that includes bot traffic overstates what the tool actually identifies for human buyers.

Six ways identified visitor data changes what a LinkedIn ad team can do

Person-level identification from LinkedIn ad traffic is not a reporting upgrade. It creates six distinct actions that were not possible from click data alone. Each shifts the team from observing to acting.

First, confirm targeting accuracy before scaling spend. When identified visitors match the ICP, the campaign's targeting is working. When they don't, the audience definition needs adjustment before budget increases. Use visitor identity to check whether the people showing up are actually the people the ad was built for. Second, prioritize follow-up by individual intent, not account name. An identified visitor who viewed the pricing page three times and checked case studies is a different conversation than a first-time blog reader from the same company. Lead scoring rules built on ICP fit (industry, company size, job title, seniority) and behavioral signals such as pages visited, time on site, and return visits produce a ranked list without manual triage.

Third, trigger CRM workflows automatically on identified visitors. When integrated with HubSpot or Salesforce, visitor identity fires customized workflows based on ICP match and buyer journey stage, without a form fill. Webhook or API delivery means the event reaches the CRM in real time, not in a morning batch export. Fourth, build retargeting audiences from actual companies and individuals, not browser cookies. You can populate LinkedIn retargeting audiences with identified visitors without relying on the third-party cookies the Insight Tag depends on. Auto-updating nurture lists defined once by criteria (seniority, industry, page visited) feed directly into LinkedIn retargeting sequences.

Fifth, feed campaign feedback into audience refinement in real time. When identified visitors cluster around a particular segment, such as mid-market SaaS companies landing on integration pages, you can take that pattern back into campaign setup as targeting intelligence. Campaigns stop running on original assumptions and start reflecting observed behavior as they run. Sixth, alert sales at the right moment, not on a weekly report. A Slack or Teams notification fired when a target account's VP visits the pricing page for the third time is a different kind of signal than a CRM dashboard reviewed on Friday. The timing of outreach (reaching a prospect while they are actively researching) is a distinct advantage over batch-export workflows. The stakes are quantifiable: 98% of B2B website visitors leave without filling out a form.

What to look for when choosing a visitor identification tool for LinkedIn ad traffic

The right tool is not necessarily the one with the highest advertised match rate. It is the tool whose identification methodology, compliance approach, and integration stack match how a B2B team actually works. Seven criteria separate tools that close the gap from tools that only claim to.

First, confirm whether the tool returns a named person with email and title, or a company name only. These are different products at different price points. RB2B identifies anonymous website visitors at the person level for U.S. traffic, and its paid plans deliver LinkedIn profiles and business email addresses, but its free tier is limited to company-level identification. Tools like Leadfeeder (Dealfront) and Albacross resolve to the company level and are designed for GDPR-friendly coverage. Leadfeeder offers broader global reach, while Albacross focuses primarily on European markets. Second, examine identity graph quality and graph size. A larger identity graph that is updated more often produces higher deterministic match rates. Ask vendors how often graph data is refreshed and what the primary identifier source is.

Third, look for journey depth, not just an identification ping. A LinkedIn profile delivered to Slack is a notification. A full event journey (page views, clicks, scroll depth, video plays, exit intent, form submissions, and the ad source) is a lead. Fourth, check CRM and ad platform integrations. Multi-touch attribution requires all marketing platforms (Google Ads, Meta, LinkedIn, HubSpot, website analytics) to feed unified data. A tool that stays in its own dashboard breaks the loop. Native integrations to check include Slack, Microsoft Teams, HubSpot, Salesforce, Attio, and direct ad platform connections for retargeting and attribution.

Fifth, ask about AI agent and bot filtering before you accept a match rate. A rate calculated against sessions that include GPTBot, ClaudeBot, and other AI crawlers overstates human identification. Require vendors to confirm that bot and AI agent sessions are excluded from the denominator. Sixth, verify compliance enforcement by region, not by trust. Person-level identification outside the U.S. requires legal review. Tools that automatically limit non-U.S. traffic to company-level resolution remove the compliance burden from you. Seventh, if your agency team manages multiple clients, demand partner accounts with domain-scoped access controls and API keys per client, not a single shared login. RB2B's Identity Graph Waterfall reaches 35–45% coverage on Pro+ plans, which provides a reference point when evaluating other vendors' claims.

From click to contact: putting the full identification workflow together

The value of person-level identification is only realized when every step from pixel to pipeline is connected. Identification alone, sitting in a dashboard, does not move deals. The full motion, however, is straightforward to describe, and each step depends on the one before it.

A LinkedIn ad fires, a visitor clicks, and the Insight Tag records the event while the Conversions API captures it server-side. The visitor identification tag then resolves the session: company-level for international traffic, person-level for U.S. visitors, with bot and AI agent sessions filtered before the record is written. An enriched record is assembled with name, title, company, LinkedIn, email, phone, seniority, and the full behavioral journey (pages, scroll depth, time on pricing page, return visits). ICP scoring runs automatically against defined criteria, and a hot-lead flag is set if the visitor matches. A webhook fires to the CRM, a Slack notification goes to the owning rep, and the visitor lands in a LinkedIn retargeting audience for continued nurture if immediate outreach is not appropriate. Finally, attribution data (which LinkedIn campaign, which ad, which targeting segment) travels with the record into the CRM, so closed-won revenue can be connected back to the specific ad exposure that started the journey.

The difference between a workflow that works and one that doesn't is usually two steps: identification without journey depth produces outreach that feels generic, and attribution without ad-source data produces pipeline reporting that cannot justify budget decisions. A visitor who clicked a LinkedIn ad, spent meaningful time on the pricing page, and returned two days later is not a cold prospect. They are a warm buyer who has already made a partial decision. The outreach message and its timing should reflect that. The point of the machinery is not that someone was identified. It is that the right person was reached at the right moment with the right context.