the paperwork
Privacy Policy
last updated · august 2026
the short version
I collect your email address to send you The Measured Life. I don't sell it, share it with advertisers, or send you anything other than what you signed up for. You can unsubscribe anytime. That's it.
01 who this is
This site — averagejoematt.com — is a personal project run by Matthew, a solo developer in Seattle. It is not a company. There are no advertisers and no investors; any affiliate arrangements are limited to the disclosed product links on the gear page.
Questions or requests: matt@averagejoematt.com
02 what I collect
When you subscribe, I collect:
- Email address — required to send The Measured Life.
- Why you're here / how you found this — optional free-text fields on the subscribe form. Helps me understand the audience. Never shared.
- Subscribe source — which page or referral link you used.
- Confirmation timestamp — when you confirmed your double opt-in.
I do not collect your name unless you volunteer it. I do not collect payment information — this newsletter is free. I do not use tracking pixels or third-party analytics in the emails.
03 how I use it
Your email address is used for exactly one thing: sending you The Measured Life, which you signed up for. I will never sell or share your email with any third party, advertiser, or sponsor.
04 where data lives
Subscriber data is stored in Amazon DynamoDB (AWS us-west-2, encrypted at rest). Email delivery is handled via Amazon SES. No subscriber data is stored in any third-party marketing platform or CRM.
05 unsubscribing
Every email includes an unsubscribe link. Click it and you're removed immediately — no re-engagement emails, no questions asked. You can also email matt@averagejoematt.com and I'll remove you within 24 hours.
After unsubscribing, your email is deleted from the subscriber list entirely.
06 cookies and tracking
This site uses no tracking cookies and no third-party analytics. There is no Google Analytics, Meta Pixel, or equivalent.
The one thing I do look at is standard server access logs — the same first-party request records every web server keeps. Once a week I aggregate them into a simple count of page views, how many distinct visitors came, and which ones returned, so I know whether anyone finds this useful. To count distinct visitors I hash the IP address in memory and immediately discard it; no raw IP, no individual visitor, and nothing about you personally is ever stored, emailed, or linked across sessions. The output is plain totals.
07 your rights
You can request a copy of any data I hold about you, ask for corrections, or request deletion. Email matt@averagejoematt.com and I'll respond within 7 days.
08 AI-generated content
The /ask and /board pages use AI (Claude by Anthropic) to generate responses based on my personal health and life data. Answers are informational only — they reflect patterns in my data and should not be treated as medical, nutritional, or professional advice.
Questions you type into /ask or /board are sent to Anthropic's API to generate a response. I do not store your questions. Anthropic's data handling is governed by their privacy policy.
09 what happens to all this if I stop
The question friends actually ask, answered in writing rather than in reassurance. Every night the whole public record is packaged into a single download at a fixed address, with a manifest that lists every file and its checksum, and a clock that measures how long it has been since the platform last heard from me. The terms below are the contract; the numbers above them are what it produced last night. Both are checkable — that is the point of writing it down.
the archive right now
- built
- reading…
- download size
- reading…
- files inside
- reading…
- edition
- reading…
- continuity
- reading…
sha-256 of the download reading…
-
P1Nothing here is for salepolicy
No reader's email address, question, or submitted finding is ever sold, rented, or handed to an advertiser, a data broker, or a sponsor. Neither is any of Matthew's own health data. There is no advertising business here to be tempted by.
-
P2The public record is rebuilt as a single download every nightmechanism
Every night the platform packages everything it already publishes — the data, the methods, the chronicle, the sealed pre-registrations — into one compressed archive at a fixed address, and overwrites yesterday's. It is not generated on request and it is not gated: it is simply there.
-
P3What may enter the archive is decided by a gate, not by handmechanism
Nothing enters because someone remembered to add it. An artefact enters only if the public site already serves it at a public address, and only if a registry classifies that address as part of the published record. Anything the registry does not classify is refused. The API portion is fetched with no credentials at all, over the same public address a reader would use, so it can contain nothing a reader could not already fetch themselves.
-
P4Every archive publishes its own inventory and checksummechanism
Alongside the archive sits a manifest listing every file in it with a size and a SHA-256, the checksum of the archive as a whole, the moment it was built, and — in the same document — a list of what was deliberately left out and why. You do not have to trust the description; you can check it.
-
P5Silence is measured out loudmechanism
The platform watches for its own silence: the number of days since the last signal from any source that only produces data when a living person is doing something. That number, and the state it puts the contract in, are published in the same place as the archive, every night, whether the news is good or not.
-
P6Ninety days of silence freezes the record and raises the alarmmechanism
At 30 days of silence the contract enters a notice state, at 60 a warning state, and at 90 the switch trips: the archive stops being overwritten and is sealed as a final, dated edition, the published state says so plainly, and the people configured as continuity contacts are told where the record is and how to check it. Any new signal from any watched source resets the clock — the switch is a measurement, and it is reversible.
-
P7What this contract does not promiselimit
It does not promise the archive outlives the bill: everything here sits in one cloud account, and if that account lapses the archive lapses with it. This edition grants no mirroring rights, so that single point of failure is not mitigated by anything written here. It does not promise a third-party mirror — no automatic copy is made to any host outside this account, and any claim that one exists would be false. It does not promise the archive is complete: audio, artwork, and the site's presentation layer are excluded, and the manifest names every exclusion. It does not promise the switch detects a person: it detects the absence of data, which a long holiday, a hardware failure, or a vendor lockout can also produce — that is why the first two thresholds notify rather than act. And it does not promise to publish anything currently private: the archive is a repackaging of the public surface, never a widening of it.
-
P8The terms are versioned, and amendments are append-onlymechanism
This contract carries a version and a dated amendment history. Clauses are added or superseded, never quietly rewritten, and the published version number tells you which edition you are reading.
10 changes
Material changes will be notified to subscribers before taking effect. The last-updated date above reflects the current version.