Short answer: Across 1.2 million identified B2B website sessions in Q2 2026, fewer than 3% involved a form submission, so 97%+ of in-market visitors stayed anonymous to the sales team. 36% of that research happened on mobile, 1 in 3 sessions came from a returning visitor, and enterprises (5,000+ employees) were the single largest identified company-size group. The takeaway: the form is not the funnel, and most B2B buying research is invisible unless you identify it.
Most published numbers about B2B website traffic come from survey panels or third-party estimates. This study is different: it is built from first-party identification data, real sessions on real B2B websites, aggregated and anonymized. We wanted to answer one question that every revenue team asks and almost no one has hard data on: what does the traffic you are not capturing actually look like?
Methodology
This study aggregates 1,209,755 identified visitor sessions across 10 B2B company websites that use Abmatic AI, recorded between April 1 and June 30, 2026. Every figure below is a pooled aggregate. No individual company, visitor, contact, or customer is identified, and any breakdown drawing on fewer than three customer sites is suppressed for anonymity. The sites span education, software, healthcare, financial services, retail, government, real estate, insurance, and construction, and skew toward North American mid-market and enterprise audiences. Treat the percentages as directional for a comparable B2B traffic mix, not as universal constants.
One deliberate choice: we do not publish a single headline "match rate." As we explain at the end, that number is meaningless without your specific traffic mix, and quoting one is the most common way visitor-ID vendors mislead buyers.
Key findings at a glance
| Metric | 2026 figure |
|---|---|
| Identified B2B sessions analyzed | 1,209,755 |
| Sessions with a form submission | Fewer than 3% (upper bound 2.5%) |
| Sessions that never submitted a form | 97%+ |
| Research happening on mobile (of known device) | 36% |
| Returning visitors (of known) | 33% (1 in 3) |
| Traffic arriving via LinkedIn | 9.2% (about 1 in 11) |
| Largest identified company-size band | 5,000+ employees (33% of known) |
| Avg unique pages viewed per identified visitor | 11 |
1. The form is not the funnel: 97%+ of identified sessions never convert a form
This is the headline. Of 1.2 million identified B2B sessions, fewer than 1 in 40 included a form submission. The other 97%+ researched your product, read your content, and in many cases checked your pricing, then left without ever raising their hand. They are not low intent. They are simply early, cautious, or doing the silent evaluation that defines modern B2B buying. If your pipeline depends on form fills, you are working from less than 3% of your actual demand.
This is the entire case for anonymous website visitor tracking: the value is not the 3% who self-identify, it is making the other 97% visible so sales can act on real interest instead of waiting for a form.
2. B2B research is mobile: 36% of identified sessions
| Device | Share of identified sessions (known device) |
|---|---|
| Desktop | 63.6% |
| Mobile | 36.0% |
| Tablet | 0.3% |
The myth that B2B is a desktop-only, 9-to-5 motion is wrong. More than a third of identified research now happens on phones, often the first touch before a buyer ever opens your site on a work laptop. If your site experience, personalization, and identification pipeline are not built for mobile, you are blurring out a third of your buyers.
3. One in three sessions is a returning visitor
Of sessions where we could classify the visitor, 33% were returning, not first-time. Repeat visits are one of the strongest, most underused buying signals in B2B. A company that comes back three times in a week is telling you something a single anonymous pageview never will. Yet most teams treat every session as net-new because they cannot connect visits to an account over time. Identification plus visit history is what turns "someone looked" into "this account is actively evaluating."
4. The company-size barbell: enterprise and small businesses lead
| Company size (employees) | Share of identified visitors with known size |
|---|---|
| 5,000+ | 32.6% |
| 1 to 49 | 23.4% |
| 200 to 999 | 15.2% |
| 50 to 199 | 14.8% |
| 1,000 to 4,999 | 13.9% |
Among identified visitors with a known employee count (about 47% of identified companies), the distribution is a barbell: large enterprises are the single biggest group, with small businesses close behind. The practical lesson is that one routing rule does not fit your traffic. Enterprise visits deserve immediate sales attention and account-based follow-up, while high-volume SMB interest is often better served by automated nurture and self-serve. Scoring by fit and behavior, not treating every identified company the same, is what makes this data usable.
Skip the manual work
Abmatic AI runs targets, sequences, ads, meetings, and attribution autonomously. One platform replaces 9 tools.
See the demo →5. Industry mix of identified B2B traffic
| Industry | Share of identified visitors with known industry |
|---|---|
| Education (higher ed, K-12, and education management combined) | 23.5% |
| Computer software | 6.5% |
| Hospital and health care | 6.1% |
| Financial services | 5.3% |
| IT and services | 3.7% |
| Retail | 3.5% |
| Real estate | 3.1% |
| Government administration | 2.7% |
This mix reflects the industries served by the sites in the sample, so read it as the composition of these B2B audiences rather than the internet at large. The broader point holds regardless of vertical: identified traffic clusters into a handful of dominant industries, and knowing yours lets you tailor messaging, segments, and sales plays to the verticals that actually show up.
6. LinkedIn is identifiable traffic
About 9.2% of identified sessions, roughly 1 in 11, arrived via LinkedIn. Social and paid social traffic is often dismissed as un-trackable brand spend. In practice it is a meaningful, identifiable slice of in-market research. If you are running LinkedIn ads and only measuring form fills from them, you are undercounting their real contribution to pipeline by a wide margin.
7. Identified visitors are deep researchers
The typical identified B2B visitor viewed about 11 unique pages (18 pageviews in total across their visits). These are not bounce-and-leave sessions. They are multi-page, multi-visit evaluations, exactly the behavior you want to catch in the consideration stage, and exactly the behavior that never shows up in a form-fill report.
Why we will not quote you a single "match rate"
The most common question in this category is "what is your match rate?" and the most honest answer is "it depends, and anyone giving you one number is selling you something." Identification rate swings hard with your traffic mix: the share that is corporate versus consumer ISP, desktop versus mobile, domestic versus international, and bot-filtered versus human. A site with heavy enterprise desktop traffic will see a very different rate than one with broad consumer-ISP mobile traffic, using the exact same technology.
Independent guidance consistently puts realistic B2B company-level identification in the range of roughly 20% to 40% of website traffic, and warns that any vendor promising 80%+ contact-level identification is almost certainly counting company matches as if they were people. We agree. That is why this study reports the shape of identified traffic rather than a vanity rate, and why our advice is to ask any vendor for the rate on your own traffic, measured transparently, before you believe a marketing number. For a full comparison of how the major tools approach this, see our guide to the best B2B visitor identification software.
What to do with this data
Three moves follow directly from the numbers:
First, stop treating the form as your funnel. Build a motion that acts on the 97% who never fill one out, through account identification, scoring, and proactive sales follow-up. Second, segment your response by fit and behavior: enterprise and high-intent returning accounts get sales, the long tail gets automation. Third, measure paid and social channels by identified-account engagement, not just form conversions, or you will keep underfunding the channels that actually drive consideration. If you want to see your own version of these numbers, book a demo and we will show you what your anonymous traffic looks like once it is identified.
Frequently Asked Questions
What percentage of B2B website visitors fill out a form?
In this study of 1.2 million identified B2B sessions, fewer than 3% included a form submission, meaning more than 97% of in-market visitors never self-identified through a form. The exact rate varies by site and offer, but the pattern, that the vast majority of buying research never touches a form, is consistent.
What is a realistic B2B website visitor identification match rate?
Independent guidance puts realistic company-level identification at roughly 20% to 40% of B2B website traffic, depending heavily on your traffic mix. Person-level identification is far lower, and any vendor claiming 80%+ contact-level matching is likely counting company matches as individuals. Always ask for the rate measured on your own traffic.
How much B2B research happens on mobile?
In this data, 36% of identified sessions with a known device were on mobile. B2B buying research is no longer desktop-only, so identification and personalization need to work on phones, not just laptops.
Are returning website visitors a good buying signal?
Yes. One in three identified sessions in this study came from a returning visitor. Repeat visits over a short window are one of the strongest account-level intent signals available, but they only become usable when you can connect visits to an account over time.
How was this study conducted?
It aggregates 1,209,755 identified visitor sessions across 10 B2B company websites using Abmatic AI between April and June 2026. All figures are pooled and anonymized, with no individual company or visitor identified and small cohorts suppressed. See the methodology section above for scope and limitations.
Can I see these numbers for my own website?
Yes. Abmatic AI identifies the companies visiting your site, scores them by fit and behavior, and shows you the share of traffic you are currently missing. You can book a demo to see your own anonymous-traffic breakdown, or explore how anonymous visitor tracking works.
