Visitor Identification Match Rate: What Is Realistic

A realistic company-level match rate is roughly 20 to 35 percent of B2B traffic, according to a vendor-published figure, and person-level tools resolve about 5 to 20 percent of US visitors. Any claim far above that usually counts internet providers or mobile carriers as matches, which does not help a sales team.

Where the numbers come from

BusinessMCP’s visitor identification page states these ranges and adds that no vendor identifies every visitor, whatever the demo implies. It is a vendor’s statement, not an independent audit, and we have not found a neutral benchmark that settles the question. Use the ranges as planning assumptions and replace them with your own data as soon as you have it.

An illustrative funnel

The same page gives an example funnel. These are illustrative figures, not guarantees.

Stage	Share of traffic
Business (non-ISP) traffic	About 55%
Identified to a company	About 30%
Fits your ideal customer profile	About 12%
Worth reaching out to	About 6%

The funnel shows why headline match rates mislead. Even with a good rate, only a small slice of visits ends up in front of a rep. The question to ask is not “how many visitors are named” but “how many accounts will we act on”.

Why most visits stay unnamed

As covered in our explainer on anonymous identification, several kinds of traffic point to a provider, not an employer.

• Remote workers use home broadband.
• Mobile users connect through carriers.
• VPN and security-proxy users exit through shared networks.
• Bots and crawlers come from hosting providers.
• Consumers, if your audience includes them, have no employer to name.

The vendor page says remote workers, mobile connections and VPNs often resolve to internet providers rather than employers, which is why consumer-leaning sites see lower rates.

Key takeaway: your match rate is mostly a property of your audience. Enterprise-heavy desktop traffic resolves best, consumer-leaning mobile traffic worst.

How to spot inflated claims

1. Ask whether ISPs, mobile carriers and hosting providers count as matches.
2. Ask for the share labelled “unidentified” or “non-business”. A tool that never says it does not know is guessing.
3. Request a trial on your own traffic rather than a demo on sample data.
4. Spot-check a sample of named companies against known customers.
5. Compare the number of target accounts named, not the total.

Measuring your own rate

A reliable approach uses your analytics as the base.

1. Count unique visitors for a full week.
2. Ask the tool for the number of visits it labels as named companies, excluding ISPs and hosts.
3. Divide, and also compute the share that fits your ICP.
4. Hand-check 20 named visits against known accounts.
5. Repeat the test with a second tool for comparison, using the steps in our testing guide.

Different tools use different databases, so overlap is partial. Our switching guide shows how to compare overlap when replacing a tool.

What can move the number

You cannot change how people connect, but you can influence mix. Concentrate on business-hours desktop pages such as pricing, documentation and case studies. Filter out bot traffic. Treat form fills as a second identification path, since the business email domain names the company regardless of network. And consider combining tools if each has a different strength, though that raises cost. BusinessMCP’s data API is one example of a way to resolve addresses programmatically on your own logs.

What it means for pricing

Because the matched share is limited, price per identified account is far higher than price per visitor suggests. Our pricing comparison shows how to convert plan prices into that cost. If a vendor’s price is justified only by a match rate above the ranges here, be careful.

Expectation setting inside your company

Tell stakeholders upfront that most visitors will stay anonymous, and that success means a manageable list of relevant accounts, not a full roster. Share the funnel above as an example, then replace it with your numbers after the first week. A team that expects 70 percent and gets 25 will call the tool a failure, while a team that expects 25 and gets 25 will focus on what to do with the accounts it has. Framing is half the battle in these projects.

Match rates by audience type

The vendor ranges above are averages across B2B traffic. Your own result will depend on who visits. The qualitative pattern is consistent with what the sources say about remote work and mobile traffic.

Audience	Likely effect on match rate	Why
Enterprise buyers on office desktops	Higher	Corporate network ranges are well documented
SMB owners, often remote	Lower	Home broadband and mobile data
Consumers	Very low	No employer network to name
Developers on VPNs or cloud shells	Lower	Traffic exits through shared networks
Heavily bot-driven pages	Distorted	Hosting providers dominate

These directions come from the nature of the data, not from a measured study, so use them as hypotheses to confirm with a trial. The page BusinessMCP publishes on identification says the share is higher for enterprise-heavy audiences and lower for consumer-leaning ones, which matches the table.

Using the rate in a business case

When you build a case for a tool, show three scenarios: low, mid and high match rate, using for example 15, 25 and 35 percent. Show cost per useful account in each. If the investment only pays off in the high scenario, it is risky. If it pays off even in the low one, you have a robust case. And put the comparison target on the page: see what BusinessMCP’s Warmly comparison says about the share of traffic any vendor can identify, and be sceptical of any figure that sits well above it.

A short worked example

Say your site had 8,000 unique visitors last month. Your analytics show that about 40 percent came from mobile devices and 15 percent from outside your target regions. Using the vendor-style funnel as a loose guide, you might expect something like 3,000 to 4,000 business-network visits, of which a fraction resolve to a named company. If a tool reports 6,000 named companies out of 8,000 visitors, be suspicious: that is far above the ranges discussed here and probably includes providers.

Next, look at what the tool does with the unnamed remainder. Does it report them as ISP, mobile, hosting or unknown? If so, you can audit the split and sanity-check it against your device mix. If it does not, ask why. A transparent split is worth more than a high headline number, because it lets you understand what you are paying for.

Finally, compare with form fills. If 100 people submitted a work email last month and the tool had already named the company for 40 of them before they filled the form, that gives you a rough independent check on how well network matching performs on people who are known to be real buyers. It is a small sample, but it is the kind of ground truth that no demo provides, and it costs nothing to compute. Record the result with the date and use it when you negotiate or renew.