How Newsletrix works
Every screen in the app is built from one raw material: the newsletters that land in your personal
tracking address. Nothing is scraped, and nothing comes from a shared archive of other people's
emails. You choose which senders to follow, and the app analyses what they actually sent you.
1
You get a tracking address
A unique inbox address, created in Get Started. Optionally tag it with a group so the data
stays separated by client or market.
2
You subscribe competitors to it
Sign that address up to the newsletters you want to watch, or forward emails you already
receive.
3
Newsletrix collects and parses
Each email is decoded, converted from HTML to text, and stripped down to subject, preheader,
links, images, CTAs, sender headers and send time.
4
AI analyses each send
A structured analysis returns tone, topics, offers, CTA quality, structure, SWOT items and
recommendations for that specific email.
5
The app aggregates it
Every screen is a different cut of the same analysed set: by brand, by time, by content type,
by link, by topic.
A longer version of this, including the full list of fields the AI returns for every newsletter and
what each plan includes, is in
How the analysis works.
Questions people ask before reading further
Estimated is not measured
Newsletrix sees competitor newsletters the way a subscriber sees them. It cannot see another
sender's real opens or clicks, and it never claims to. Anything labelled est. is a model
estimate built from the content of the email itself. Use it to rank emails against each other,
not as a number to report to a client as fact.
Does tracking a competitor tell them anything?
You appear as an ordinary subscriber, because you are one. The sender sees an address on their
list. Newsletrix does not open, click or interact on your behalf, so no engagement signal is sent
back beyond the subscription itself.
How much data do I need before the screens mean anything?
A single newsletter already fills the detail view, the CTA breakdown and the SWOT items for that
send. Screens that compare, rank or look for trends need repetition: roughly five sends per brand
before rankings stop moving with every new email, and about a month before send-time and cadence
patterns are worth acting on.
Can I separate clients or markets?
Yes. Tracking addresses can carry a group label, and the workspace selector at the top of the
sidebar scopes the app to one of them. See
Settings and workspaces.
Where do I report something that looks wrong?
Use the feedback option in the app, or the contact page. If a specific
newsletter was parsed badly, note its subject line and sender, since that is what makes it
reproducible.