The Personalization Proof
How might a publisher prove, with data, that a personalized paper deck changes what a person retains and returns to — without ever surveilling the reader?
In your first hour
- Pull the two live data contracts below. Profile them: 104 card rooms in one, every recorded reading with its quality scores (MOS, similarity, pass/flag verdicts) in the other. Real production data, publicly served.
- Write down the three things a paper press could measure through QR scans — and one thing it must refuse to measure even though it could. That refusal list is graded work here.
- Sketch the KPI tree from “a card is scanned” to “a person was helped” — and mark every link in the chain you don’t believe yet.
Materials on the table
- rooms.json — the room ledger
- audio.json — readings + quality gates
- /evidence/ — the research shelf
- /grief/ — the 52 open rooms
The season version: a measurement framework for the paper-digital learning loop, an analysis of real scan data, and the privacy-by-design specification that decides what this press never collects. Ask Michelle for the full brief.