# Identification rate: why the figures vary so much

> Providers quote very different identification rates. Seven factors explain the range. How to measure the figure for your own website.

Quelle: https://whovisit.app/en/blog/visitor-identification-rate-why-figures-vary
Veröffentlicht: 2026-11-07
Herausgeber: Whovisit (ARAVILO, Innsbruck)

Anyone comparing visitor identification providers finds rates that lie far apart. Sometimes it sounds as if you see almost everyone, sometimes as if you see only a small part. In this article we deliberately give no figure and quote no one else's figures either. The reason: a single rate says almost nothing without context. It is more important to understand what it depends on.

## What "identified" actually means

With assignment via the network, the network owner is looked up for every visit and assigned to a class:

| Class | Meaning | Result |
|---|---|---|
| Company | Network owner with legal form, or an institution | Visit is assigned to a company |
| Internet provider | Access network, private, home office, mobile | No company recognisable |
| Data centre | Cloud, host, VPN, relays, scanners | No company recognisable |
| Without legal form | Could be a human being | Not analysed |

Identified here means: there is a network owner with a legal form, or an institution. Whether that is the company you mean is a second question.

How the lookup in the registry works is shown in [Reverse IP lookup explained simply](https://whovisit.app/blog/reverse-ip-lookup-erklaert). An overview of all the ways to identify companies behind visits is given in [Which companies visit my website?](https://whovisit.app/blog/welche-firmen-besuchen-meine-website).

## Seven factors behind the fluctuations

1. **Target group.** On a B2B site for industrial customers, more visits come from company networks than on a guide for private individuals.
2. **Place of work.** Someone sitting in the office comes via the company network. Someone working from home comes via their internet provider's network.
3. **End device.** Visits via mobile networks run in the mobile operator's network, even if the phone belongs to a managing director.
4. **VPN and relays.** Company VPNs can make a company visible, private VPN services and relays hide it.
5. **Corporate networks and service providers.** Here the registry often lists a head office or an IT service provider instead of the company you want to identify.
6. **IPv6 at the host.** If the analysis runs on a server that sees visits only via a provider's IPv6 addresses, companies are lost. That is why the technical environment of the service matters.
7. **Consent rate in the banner.** If only some visitors consent, only that part can be identified at all. Good banner design changes your figures more than any technology behind it.

How home office, mobile networks and VPN work in detail is explained in [Home office, VPN and mobile networks: who you cannot see](https://whovisit.app/blog/homeoffice-vpn-mobilfunk-besuchererkennung).

## Why a blanket rate says nothing

First, the reference base differs. A rate can refer to visits, to page views, to sessions or to individual visitors. Depending on which, the same website looks different.

Second, misassignments are counted too. Anyone who counts every network owner as a hit, including hosts or internet providers, easily reaches a high rate that is worthless for sales. A lower figure with cleanly separated classes can be more useful.

Third, the target group is a lever that has nothing to do with the quality of the tool. The same technology delivers very different results on two different websites.

What to look for when comparing providers instead of a rate is covered in [Comparing visitor identification: 12 criteria for choosing](https://whovisit.app/blog/besuchererkennung-vergleich-kriterien).

## How to measure it yourself

You do not need anyone else's figure if you get your own. Set aside a month for this and proceed as follows:

1. **Determine the share of company networks.** Divide the page views from company networks by all analysed page views of the last 30 days. This is not an identification rate but a value for your website.
2. **Separate by channel.** Calculate the value separately for paid ads, search engine, direct visits and newsletter. That way you see which channels bring companies at all.
3. **Test visit from the office and from home.** Open your site once from the company network and once via your private connection. The first should appear, the second should not. This shows you the limit vividly.
4. **Check a sample by hand.** Take 20 identified companies and check whether the entry matches a business that is actually plausible as a visitor. If many hits are wrong, you know the error side, not just the hit side.

If no company appears on the test visit from the office, the guide [Why do I not see any companies?](https://whovisit.app/hilfe/warum-sehe-ich-keine-firmen) helps.

In Whovisit, the distribution is shown in the report of each website below the key figures, in the section "Where the page views came from": company networks, private connections and mobile, data centres and VPN, without legal form and not determinable. Below that is the share of company networks. This is expressly not an identification rate.

## Frequently asked questions

### Is there a good rate I can use as a guide?

No. Rather compare your own figure over time and by channel.

### Can I improve the figure?

Yes, mainly through the banner, the script in the right place and the analysis by channel.