Your GA4 traffic chart is climbing. Your inquiry count is flat. You look at those two lines side by side and think something must be broken, because more visitors should mean more conversations. It’s not a contradiction. It’s an exact description of what happens to the companies that came to your site and left without a trace.
Ask yourself one question: if 40 companies that perfectly match your ICP landed on your site today, how many of them could you name? And how many would you even know about? For most B2B sites, the answer to both questions is: almost none. The median B2B conversion rate is 2.9% (First Page Sage, 2026), which means more than 97 out of every 100 visitors leave without giving you anything.
This article will show you how to price your own gap on your own data in 10 minutes, and give you an honest range of what can realistically be recovered. Without the promise that you’ll “identify every visitor,” because that’s a promise the math won’t let anyone keep.
Key Takeaways – Most companies that reach a B2B site leave unidentified. That’s not a measurement glitch, it’s the way buying works today. – Buyers now complete most of the buying journey before they ever reach out to a vendor. They spend just 17% of their time meeting with any single vendor (Gartner, 2024). – In a study on a large Polish B2B site, the free identification layer delivered 0% effectiveness across three independent approaches, and traffic was 55% home connections and 37% mobile. – Lost pipeline can be quantified on your own numbers and partially recovered. You can’t recover all of it.
What anonymous B2B traffic actually is
Anonymous B2B traffic is the companies that visit your site and leave without leaving any identifier: no form, no email, no phone number. This isn’t a fringe case. It’s the dominant scenario. If the median conversion rate is 2.9% (First Page Sage, 2026), and 55% of users abandon a form before submitting it (Surface Labs, 2025), then the rule is simple: whoever wasn’t going to reveal themselves, won’t.
Why does this happen? Because that’s how companies buy today. Buyers spend just 17% of the entire buying journey meeting with any vendor (Gartner, 2024), and 67% of them flat-out prefer a rep-free experience (Gartner, 2026). Someone reading your article at three in the afternoon, comparing you against three competitors, and coming back in two weeks is doing exactly what they’re supposed to do. They’re just doing it invisibly.
Think of it like a shop with a window facing a busy street. Hundreds of people glance through the glass, look, size you up, and some will come back. You don’t have a list of names of the ones who paused. You only have the ones who walked in and came up to the counter. Anonymous traffic is that entire street in front of the window.
Why most of this traffic can’t be identified
Let’s start with a sentence you won’t hear from anyone selling identification: most of your visitors physically cannot be tied to a company, no matter how much you pay for it. This isn’t a tool problem. It’s a traffic-structure problem.
In a study on a large Polish B2B site, we tested this on our own data. The free identification layer delivered 0% effectiveness across three independent approaches. Zero. And when we broke the traffic down by source, the picture became obvious: 55% was home connections and 37% mobile networks. The remaining 8% was hosting and data-center addresses (4%), bots (3%) and traffic from abroad (1%). Traffic from corporate networks, the only kind an IP lookup can tie to a company, was practically zero.
Where does this structure come from? From how people work today. Nearly 80% of workers whose jobs can be done remotely work hybrid or fully remote (Robert Half, 2025), and 42% of B2B buyers research on mobile devices (Digital Commerce 360 / Google, 2025). A director reading your offer from a phone on a train, or from home wifi in the evening, has no corporate address at that moment that anything could be read from. VPNs add another layer of cover.
What does this mean in practice? That there’s a hard, structural ceiling on what’s identifiable at all. Vendors love to talk about “30 to 65% of B2B traffic” at the company level (that’s vendor data, not independent measurement), but those numbers quietly assume the visitor is sitting behind a corporate connection. In our sample, practically none were. The rest is the street in front of the window, and no camera will capture it.
Price your own gap in 10 minutes
The answer is simple: you don’t need any tool to estimate how much pipeline you’re leaking. You need five numbers you already have, and ten minutes. The logic is one line: traffic, times the share of traffic from corporate networks, times ICP fit, times deal value, gives you lost pipeline.
Let’s walk through it step by step. Take your monthly session count from GA4. Cut the traffic that structurally can’t be tied to a company (in our sample that was practically all of it, so check your own share before you assume anything). From what’s left, deduplicate down to unique organizations. Filter to companies that fit your ICP. Subtract the ones already in your CRM. Multiply by your average contract value. The number that comes out is your monthly exposure: the pipeline walking past you anonymously.
Example: 4,000 sessions, how many ICP companies?
Let’s say you have 4,000 sessions a month and, unlike in our sample, 5% of them come from corporate networks. That share is an assumption for this example, not a benchmark.
- Corporate-network traffic, 5% (assumed): ~200 sessions.
- After deduplicating to unique organizations: ~75 companies.
- ICP fit, say 20%: ~15 ICP companies.
- Net new, not in CRM: ~9 companies a month.
- Average contract value of EUR 10,000: that’s EUR 90,000 of potential pipeline exposure per month walking out the door without a trace.
Stop on that last number. You won’t lose 90 thousand times twelve, because you wouldn’t recover all of that pipeline anyway. What changes is that you stop guessing. You have a concrete number instead of a hunch that “something’s leaking.” That’s the difference between a budget built on fear and a budget built on arithmetic.
Plug in your own numbers. If you have 12,000 sessions and contracts worth EUR 30,000, or a much smaller corporate-network share, the result will look completely different, and only then will you see whether this problem deserves any attention at your company.
What can realistically be recovered
The honest answer is: some, not all, and less than the market promises. The exposure from the previous section is a ceiling, not a forecast. Of those ~9 companies a month, a real, manual process will recover not all of them, just a handful of the ones that were actually in a buying stage rather than passing through by chance.
A realistic range? From the identifiable slice, careful work can turn about 1 to 2 companies a month into an actual conversation. At a EUR 10,000 contract, that’s EUR 10,000 to 20,000 of real pipeline per month, which over a year is an amount that for many B2B companies justifies the whole effort on its own. This isn’t “recover every visitor.” This is “recover the part you can honestly reach.”
There’s one more thing. The ceiling on this work is set not by budget, but by the traffic structure you calculated above. You can improve your hit rate within the corporate-network slice, if your traffic has one. You won’t move the home and mobile traffic, because there’s nothing to identify there. Anyone who claims otherwise is selling you a number they won’t hit.
What an “identified” visitor actually gives you
Let’s say you go into paid identification and start getting company names from that corporate-network slice. Before you celebrate, one question: is a company name the same as a contact you can write to? And is it the same as buying intent? The answer is no and no.
A company name tells you that someone from that organization was on the site. It doesn’t tell you who: the CEO or an intern, a decision-maker or an analyst doing research for a thesis. It doesn’t tell you whether it’s a buying signal or a coincidence. Between “company X was on the site” and “I have someone to talk to and something to talk about” lies all the work that identification alone doesn’t do. Part of that work is lead routing: deciding who owns the account and acts on it.
It’s worth setting this against how buyers decide today. 94% of buying groups have their preferred vendors ranked before the first contact, and in 77% of cases they buy from that early favorite (6sense, 2025). On top of that, in 83% of situations it’s the buyer who initiates contact, not the rep (same source). A company name alone doesn’t tell you where on that list you stand. That sets a realistic expectation: identification is the start of the work, not the end of it.
What good looks like
A healthy state isn’t knowing every visitor. It’s that you’ve stopped demanding it and started counting. You’ll recognize it by three things.
First, you have a number. You know how many ICP companies pass by you anonymously each month, and you know that number has a structural ceiling. Second, you work on the slice you can reach, instead of burning budget on a promise of 100% coverage. Third, you treat every identified company as the start of a conversation to be had, not a lead ready to close. When those three things are in place, you stop chasing the ghost of full identification and start recovering what’s real. You can check that number alongside the rest of your funnel in a B2B marketing audit.
FAQ
How do I know who visits my website? To a limited degree, and only for part of your traffic. In our sample, home connections (55%) and mobile networks (37%) made up most of the traffic, and traffic from corporate networks was practically zero, so an IP lookup had almost nothing to work with. Where corporate traffic exists, a company name can sometimes be established, but the specific person usually can’t.
Can you identify anonymous website visitors? Partially. Companies on corporate networks are sometimes identifiable, but not individuals, and not mobile or home traffic. In a study on a large Polish B2B site, the free identification layer delivered 0% across three independent approaches.
How much B2B traffic is anonymous? The vast majority. At a median conversion rate of 2.9% (First Page Sage, 2026), more than 97 out of every 100 visitors leave no identifier at all.
What is anonymous B2B traffic worth? As much as your own arithmetic says: sessions times the share of corporate traffic, times ICP fit, times contract value. For 4,000 sessions, an assumed 5% corporate-network share and EUR 10,000 contracts, the exposure is on the order of EUR 90,000 a month, of which you’ll realistically recover some, not all.
Is there a free way to see who’s been on the site? The free identification layers in our study delivered 0% effectiveness. Instead of hunting for a free tool, it’s cheaper and more honest to first calculate whether this problem is even worth a budget at your company.
Before you spend anything on identification, calculate your number
Three things to remember. Anonymous B2B traffic is the norm, not a glitch, because that’s how companies buy today. Lost pipeline can be calculated on your own numbers in 10 minutes. You’ll recover some, not all, and anyone who promises more is selling you a number they won’t hit.
You now have the framework. The question is: do you want to keep guessing, or see how much pipeline you’re really leaking, on your own data?
→ Pipeline Diagnostic (27 points): see how much pipeline you’re really leaking.
→ Prefer to talk specifics? Book an audit.
Next in the series: what visitor identification can and can’t actually do →