Visitor Identification Match Rates: The Real Numbers
Every visitor identification vendor leads with a match rate, and almost none of them are measuring the same thing. One tool's "98 percent accuracy" describes a verified subset of its database. Another's "20 percent" honestly describes person-level resolution of raw traffic. Comparing the two numbers directly is like comparing one restaurant's hygiene rating to another's Yelp stars. This guide lays out the published numbers as of July 2026, explains why they diverge so wildly, and gives you a two-week protocol to measure the only match rate that matters: yours.
What "Match Rate" Actually Means (Three Different Things)
Vendors use one term for three metrics. Coverage: what share of total visitors get identified at all. Resolution level: whether an identification means a company name or an actual person. Accuracy: of the identifications made, how many are correct and complete. A tool can score high on one and badly on another, and the marketing will always quote the flattering one. The useful mental model is coverage multiplied by accuracy at the resolution level you need: a tool that's 98 percent accurate on 15 percent of traffic identifies fewer real people than one that's 85 percent accurate on 60 percent.
The Published Numbers, Side by Side
| Tool / category | Published or reported figures (July 2026) | What it measures |
|---|---|---|
| Company-level tools (Leadfeeder, Snitcher, Leadinfo) | Commonly 40–70% of B2B traffic | Company named per session, office networks strongest |
| RB2B (standard plans) | 5–20% of US traffic | Person-level, US only |
| RB2B (Pro+ premium waterfall) | 35–45% of US traffic | Person-level, US only |
| Beam | 60–80% of visitors on average | Person-level, layered identification with LLM fallback enrichment |
| B2B contact data industry broadly | 80–90% accuracy in operational tests | Accuracy of records, not coverage of traffic |
Every row is a claim by the vendor or by reviewers, measured on different traffic with different definitions. Including ours: Beam's 60 to 80 percent is our published average, B2B traffic skews to the top of it, and you should verify it on your own site exactly like every other number here.
Why the Numbers Swing So Hard
Four factors drive most of the variance. Traffic geography: US traffic matches identity graphs far better, which is why some vendors restrict person-level identification to it. Network type: office networks resolve cleanly; remote workers, VPNs, and mobile carrier traffic resolve poorly, and the shift to hybrid work permanently lowered ceilings across the industry. Audience type: B2B professional audiences appear in identity graphs at much higher rates than anonymous consumer traffic. And method: single-source reverse IP tools cap out early, while layered approaches (IP plus device signals plus identity graphs plus enrichment fallbacks) stack coverage, which is the design reason behind Beam's multi-layer pipeline, explained in our person-level identification guide.
How to Measure Your Own Match Rate in Two Weeks
Vendor benchmarks end arguments; your own data starts pipelines. The protocol is simple. Install two or three tools with free tiers side by side (each is one script tag, and they don't conflict; our roundup lists which have free plans). Run them for two full weeks to smooth out weekday and campaign noise. Then compute three numbers per tool: coverage (identified visitors divided by total unique visitors from your analytics), quality (spot-check 20 identified records; how many name the right person at the right company), and actionability (how many identifications came with enough context, like a role and an active social profile, that you could actually reach out today).
That third number is the one to weight, because it's the one connected to revenue. A name without a channel is trivia. This is also where tools diverge most: Beam attaches matched social profiles and a drafted opener to each identification, so an identified visitor is one click from contacted, a workflow difference covered in our pillar guide.
Red Flags When Reading Vendor Claims
Be suspicious of accuracy numbers with no coverage number attached, of "up to" phrasing doing heavy lifting, of benchmarks measured only on the vendor's own site traffic, and of any tool that won't let you test before paying. The free-tier test is the great equalizer in this category: a vendor confident in its match rate will let your traffic prove it. Ours does; the setup takes about 30 minutes.
FAQ
What is a good match rate for website visitor identification? Depends on resolution level and traffic. For company-level identification of B2B traffic, 40 to 70 percent is the commonly reported band. For person-level, published figures range from 5 to 20 percent (RB2B standard, US only) to Beam's reported 60 to 80 percent average. Judge any number against your own two-week test.
What percentage of website visitors can be identified by name? It varies by vendor method and your audience. US-based, B2B-professional traffic identifies at the top of every tool's range; international, consumer, and VPN-heavy traffic at the bottom.
Why do vendors report such different match rates? Because they measure different things (coverage vs accuracy vs resolution level) on different traffic. There's no standardized benchmark in this category, which is why side-by-side free-tier testing beats spec-sheet comparison.
How do I test a visitor identification tool's match rate? Install it alongside your analytics for two weeks, divide identified visitors by total uniques for coverage, spot-check 20 records for quality, and count how many identifications you could act on the same day.
every vendor has a number. your traffic has the answer. run the two-week test, starting free. get started free →