Most “match rate” conversations collapse two different things into one. The first is company-level identification, which tells you “someone from Acme visited.” The second is person-level identification, which tells you Sarah Chen, VP Marketing at Acme, visited. The first is an easier problem and produces higher headline numbers. The second is the one that actually drives outreach.
I am George, founder of Leadpipe. Our identity graph operates at the person level because that is what the sales motion actually needs. This post is the 2026 read on how match rate varies by industry, geography, and traffic composition, what 30-40%+ on US B2B actually means in context, and how to read your own number without getting fooled by vendor headline figures.
The honest baseline
Leadpipe’s average person-level match rate on US B2B traffic is 30-40%+, depending on traffic quality. That is the number I publish with confidence across the customer base. It is not 50%. It is not 70%. Tools that claim higher numbers are usually measuring something else, most often company-level rather than person-level matches.
Some context on why the number lands where it does:
- It is person-level, not company-level. Resolving the named individual is a strictly harder problem than resolving the organization.
- It is deterministic, not probabilistic. We require verified linkages, not statistical inference. Probabilistic graphs report higher headline numbers and lower accuracy in independent tests (8.7/10 for Leadpipe vs 5.2/10 RB2B and 4.0/10 Warmly).
- It is US B2B-specific. EU/UK defaults to company-level only. International coverage drops sharply because identity graph depth is highest in the US.
- It is the average across all traffic to the site, including pages and channels where match rate is low. The pricing-page rate is much higher than the homepage rate. See below.
Anyone who tells you a single industry-bucketed match rate without those caveats is selling, not measuring.
Why match rate varies
Match rate is not one number. It is a function of at least five factors that move independently.
| Factor | What raises match rate | What lowers it |
|---|---|---|
| Geography | US traffic | International, especially mobile-heavy regions |
| Audience type | B2B (corporate networks) | B2C (residential, mobile) |
| Device mix | Desktop | Mobile-only |
| Traffic source | Direct, organic, branded | Programmatic display, low-quality social |
| Page intent | Pricing, comparison, integrations | Top-of-funnel blog |
Two sites in the same industry with identical traffic volumes can produce match rates that differ by 15 points if their geographic and device mix differ. The industry label alone does not predict the rate.
Person-level vs company-level
Every identity vendor reports a match rate. They do not all measure the same thing.
| Approach | What it returns | Typical match rate range |
|---|---|---|
| Company-level (reverse IP, firmographic) | “Someone from Acme visited” | 50-70% on mixed remote traffic |
| Probabilistic person-level | ”Probably Sarah Chen” | 5-20%, accuracy varies |
| Deterministic person-level | ”Sarah Chen, verified” | 30-40%+ on US B2B |
Company-level match rates are consistently 1.5-3x higher than person-level. The work is easier. You only need to resolve a visitor’s organization, often via reverse IP lookup, and you are done. Person-level requires stitching device, cookie, and behavioral signals to a named individual with verified contact data.
The practical problem: you cannot email “someone from Acme.” You need Sarah. A 60% company-level match rate that collapses to 22% person-level looks very different in the sales workflow than in the marketing slide. For the long-form version of this argument, see person-level vs company-level visitor identification.
Why industries cluster
Two patterns repeat. They are structural, not vendor-specific.
Industries that sell to digitally-active US professionals cluster at the top. SaaS, fintech, B2B e-commerce, cybersecurity, marketing agencies. Their buyers browse from corporate networks, on desktops, with predictable identity-signal density.
Industries with heavy international footprint, institutional traffic, or traditionally-offline audiences cluster lower. Manufacturing, healthcare IT (where institutional networks dominate), edtech (where school district NAT obscures identity), legal tech (where consumer-facing pages pull mixed traffic). These industries also tend to have higher mobile share, which compounds the penalty.
Within an industry, individual sites can be far above or below the bucket median depending on their specific traffic mix. The industry bucket is a starting point for calibration, not a target.
Mobile match rate penalty
Every industry has a mobile match rate that is lower than its desktop rate. The gap varies but the direction is consistent.
The structural reasons:
- Mobile devices share less identity signal density than desktops. Cookies are more aggressively partitioned. App-to-web matching is harder. Cross-device stitching from mobile to desktop is the hardest leg of the whole stitching problem.
- IP-based resolution collapses on mobile. Carrier-grade NAT means thousands of mobile users share the same IP range. Reverse IP lookup is useless. A pure-IP tool is effectively blind on mobile.
- Deterministic cookie-and-signal matching degrades on mobile but does not collapse the way IP resolution does. The relative penalty for an identity-graph approach is smaller than for a reverse-IP approach.
If your audience is heavily mobile and your tool is IP-based, you are effectively blind on roughly half your traffic. See our mobile visitor identification guide for why this matters when comparing tools.
International traffic penalty
Identity graph depth is highest in the US. This is not vendor-specific. It reflects data availability, regulatory posture, and the share of the open web that produces consented identity signal.
Directional penalties versus US visitors:
| Region | Person-level coverage |
|---|---|
| United States | Baseline |
| Canada | Slight penalty |
| United Kingdom | Larger penalty, GDPR defaults reduce coverage |
| Western Europe (excluding UK) | Larger penalty, similar reasoning |
| Australia | Moderate penalty |
| Rest of World | Significant penalty |
If your site receives significant traffic from regions below the first two rows, your total-visitor match rate will be lower than the US-only benchmark even though your US slice performance is fine. Report those rates separately when comparing benchmarks.
For European visitors specifically, GDPR-compliant visitor identification defaults to company-level only. Person-level requires affirmative consent. That is a different mechanism than coverage limitation, but it has a similar effect on the rate you observe.
Page-type variation within the same site
The same site shows very different match rates depending on which page the visitor is on. High-intent pages over-index because the audience self-selects into professional, research-driven behavior.
| Page type | Direction relative to site average |
|---|---|
| Pricing | Strongly above average |
| Product / features | Above average |
| Comparison / alternatives | Above average |
| Integrations | Above average |
| Case studies | Slightly above average |
| Blog (top-of-funnel) | Below average |
| Homepage | Mixed, depends on traffic source |
Pricing and comparison pages consistently match well above the site average. Top-of-funnel blog traffic sits below. This is one reason raw site-level averages can be misleading. Your pricing page is doing better than your homepage number suggests, and your pricing-page rate is the one that pays.
Implications for the reader
Benchmark to your traffic, not to a marketed number. A 28% person-level match rate is above average for a manufacturing site with a 40% mobile share. The same rate is below average for a US SaaS company with desktop-heavy pricing-page traffic. The category label alone does not predict the rate.
Separate company-level from person-level when vendor-shopping. If a vendor quotes 60% match rate, ask three questions. Person-level or company-level. US or global. Averaged across customers or quoted from one. The answers usually collapse the number to something comparable to a realistic person-level benchmark.
Check your traffic composition before blaming the tool. If your match rate is below the industry benchmark, the cause is almost always one of: heavy international traffic, mobile-heavy audience, or high bot share contaminating the denominator. See our bot traffic study for why bot share matters more than most teams realize.
Measure the high-intent pages separately. Pricing, comparison, and integration page match rates are the ones that translate to pipeline. A low sitewide rate can mask a strong pricing-page rate, which is the number that pays. Run the pricing-page workflow on the high-intent slice, not on aggregate.
Watch the trend, not the absolute number. Identity graphs refresh continuously. Your match rate today is a point-in-time view. Track it monthly and look at the slope. A flat trend with a 32% number is healthier than a downward trend with a 38% number.
A note on inflated headline numbers
Across the visitor-identification category, you will see vendors quote numbers that look much higher than 30-40%+. A few common ways those numbers get inflated:
- Company-level masquerading as person-level. “Match rate” without a person/company qualifier is almost always company-level if the headline is above 50%.
- Probabilistic confidence threshold. A tool can boost reported match rate by lowering its confidence threshold. The cost is accuracy. See the independent test for what that costs.
- Cherry-picked customers. “Up to 50%” usually means the best-performing single customer with the cleanest US desktop B2B traffic. The customer-base average is much lower.
- Sessions vs unique visitors. Reporting matches per session inflates the rate because returning visitors get counted multiple times.
When in doubt, ask the vendor for the average match rate across their full customer base, with person-level required, US-only required, and unique-visitor denominator required. The number will land in a believable range, or the conversation ends quickly.
Limitations
- Single-vendor data. Match rates reflect Leadpipe’s identity graph. Another vendor with different data sources will produce different rates, though the relative industry ordering tends to be consistent across vendors that build deterministic graphs.
- Site-level reporting. Match rate is computed per site, not pooled. Pooled rates are dominated by the largest sites and over-represent their traffic mix.
- Quarterly snapshot. Identity graphs refresh continuously. A 2026 number is a point-in-time view. Expect drift over quarters.
- No accuracy axis. Match rate is one dimension. Accuracy is a separate dimension. A tool with a higher match rate and lower accuracy delivers worse outreach outcomes than a tool with a slightly lower match rate and higher accuracy. Measure both.
Leadpipe identifies 30-40%+ of your US B2B visitors with full contact data on the Pro plan at $147/mo. No credit card to start the 500-lead trial. Start identifying visitors →
