Methodology

Counts on this page are read from our database when you load it. As of 2026-11-10 22:22 UTC.

What is in the database

Not counted as people: 24,286 channels, shows, organizations and accounts, 4,185 of them with a published contact address.

Who counts as a person

We count a contact as a person only when its name is a person's name. Most listings on YouTube, Spotify, podcast directories, Twitch and SoundCloud are channels or shows, so there we require positive evidence: a recognised given name, or a first and last name that is not the show's title. Listings without that evidence are counted as channels, organizations or accounts, not people. This errs toward undercounting: some channels run by one person are counted as channels.

Where a contact comes from

Our crawler reads the pages an outlet publishes about its own people: mastheads, staff pages, team pages and author pages, and the author sitemaps many outlets provide. We keep a person only if their name appears on that page. For every person we record the exact page we found them on and the time we read it. 37,307 of the 39,199 people in the database have that source page recorded.

Some contacts come from public platform profiles and feeds instead, for example a podcast's feed or a creator's channel description. Those are labelled as such.

How each email address is labelled

Every address carries one of these labels, and the label is shown with it:

1. Published on the outlet's own page: the address was on a page the outlet publishes when we read it, and we link to that page. For addresses we read since 26 September 2026 we keep the words around the address as proof (2,823 so far). When we re-read the pages behind earlier addresses on 26 September 2026, 2,197 of 2,939 were in the page text and 550 were only in the page's code, such as an email link. 192 were not on the page we had recorded. We checked each of those: 41 had been recorded against the outlet's home page, a feed or an index page, and when we read the right page 26 were in its text and 15 only in an email link; 13 are shared or role mailboxes, such as corrections@, that should not have been labelled as a person's; 77 are addresses we could not find on the outlet's pages at all, most of them recorded by our extraction software from a page that, unchanged since we read it, does not contain them (we now label these unverified and never show or use them); 3 were on the page and our re-read missed them; and for 58 we cannot yet tell, because the page has changed or could not be re-read. We send no notice to an address unless it is proven to be on the page, and we are reviewing all of these.
2. Published on a profile or feed: the address appears on a public profile, channel page or feed. On a personal profile or newsletter the person published it. On a podcast feed or channel page it is the contact address the show publishes.
3. Never published: we have no address the person or their outlet published. An older part of our system guessed addresses from name patterns (first.last@outlet). We keep those guesses only so we never mistake them for real ones. We never show a guess as an address, and it can never be marked verified without a real check.

Addresses a site has hidden

We do not decode obfuscated email addresses. When a site hides addresses with Cloudflare's email protection or a similar tool, we treat that as the publisher asking not to be harvested, and we do not record the hidden address.

Verification

"Verified" means one thing: a real message was delivered to the address and did not bounce. We show the date it happened. It is not an unexplained badge.

Checks that do not send anything cannot prove a mailbox exists, so we do not call them verification. We run them anyway to remove addresses that cannot work: malformed addresses, domains that do not exist or accept no mail, parked domains and disposable-address services. Today they have removed 27 addresses. 4,647 addresses pass them, which means only that the domain accepts mail. The mailbox itself is still unknown.

The message we send is the notice the law requires: a short email telling each person that we hold their details, where we found them, and how to be removed. If it is delivered and does not bounce within 72 hours, the address is marked verified with that date. If it bounces because the mailbox does not exist, the address is marked undeliverable. If someone replies that the person has left the outlet, we mark that too.

No address is marked verified yet. That is accurate: we have not yet sent the notice that verifies them (see "Verification" below).

How the crawler behaves

The crawler identifies itself as NewzWyreBot/1.0 (+https://newzwyre.com/crawler). It reads robots.txt first and follows it. It does not crawl a site whose robots.txt it cannot read, and it does not crawl a site that sets Content-Signal: ai-input=no. It fetches one page at a time per site, at least 10 seconds apart, or slower if the site's Crawl-delay asks for more. A site that asks for more than 60 seconds between requests is not crawled. It does not log in, submit forms, or read pages behind a paywall. Details: newzwyre.com/crawler.

Opting out

One click removes a person, or a whole domain, and it takes effect immediately. Every email we send carries a one-click unsubscribe link.

Chronic Internet LLC · P.O. Box 510072, Kealia, HI 96751, USA · Privacy notice