How NewsletterFIT collects and labels data
Last updated September 29, 2026
NewsletterFIT indexes public newsletters and their posts, then connects them to the topics, companies, people and sponsors they mention. Every number on a profile carries a label saying where it came from, so you can decide how much weight to give it.
Sources
- Public publication pages, archives, RSS feeds and sitemaps on Substack, beehiiv and custom domains.
- Public discovery surfaces such as platform category listings and newsletters' own recommendations of other newsletters.
- Links between publications, used to find newsletters we don't track yet.
We follow each source's robots rules and terms, and we never access paywalled or private content. We do not have publishers' private metrics such as open rates, click rates or ad pricing.
What the labels mean
Shown publicly
Subscriber figures are what the platform or the publication's own site displays, for example "50K+ subscribers" on Substack. They are often rounded and are not verified counts. When a publication shows no figure, or the figure looks unreliable, we show a dash instead of estimating one.
Observed
Things we count ourselves from the posts we have indexed: when the last post went out, how many posts appeared in the last 90 days, and the publishing cadence derived from that count.
Inferred
Topics and "likely audience" are classified from the text of recent posts by a language model. They describe what a newsletter writes about and who it appears to be written for. They are not reader demographics.
Detected
Sponsors are detected in posts that carry an explicit sponsorship marker, such as "Sponsored by" or "Presented by", and the brand name is checked against the post text. Only detections above a confidence threshold are shown. Detection can miss placements that are not labelled, or misread one, so treat sponsor lists as a starting point for research. Sponsor pages list every newsletter and post where a brand was detected in the last 12 months, with the sponsorship line quoted from the post.
Momentum
Momentum compares a publication's publicly shown subscriber figure between two snapshots taken on different dates. We don't show it when the starting figure is too small or the change is implausibly large, because rounded public figures make those swings misleading. The rising newsletters page applies stricter rules on top of this, listed on the page.
Freshness
Active publications are re-checked on a rolling schedule, usually at least daily, and new posts are classified within hours to a couple of days depending on queue load. Profiles show the date they were last refreshed, and the topic and directory pages are rebuilt every hour.
Corrections and removal
If something about your newsletter is wrong, or you would like a public profile removed, email support@newsletterfit.com from an address connected to the publication and we will fix it.
Browse the newsletter directory or all topics.