How Flick research earns publication
Flick does not turn a research idea into a result. A protocol becomes a published study only after the observations exist, quality checks pass, limitations are written and the underlying evidence can be inspected without exposing private email.
Email Exit Gap Index
The Index separates three evidence tiers: pending, machine capability and observed outcome. A sender receives no honor grade from a header alone. A full observation records the standard unsubscribe mechanism, human friction and whether messages from the tested subscription stop during the declared follow-up window.
- Use a dedicated research inbox subscribed directly to the named sender.
- Preserve the message headers and the exact unsubscribe path.
- Record clicks, login walls, preference steps and confirmation.
- Recheck the same subscription at the pre-registered time points.
- Publish only evidence that passes the provenance guard; private user data never enters a sender page.
The current live Exit Gap Index remains explicit about which senders are graded, capability-only or pending.
Swipe-versus-click research
A public result requires participants who are not presented as representative of all email users, matched task decks, randomized interface order, recorded completion time, correction rate and effort, plus a disclosed analysis plan. A small internal staff run may calibrate the instrument but is never marketed as an independent finding.
Multi-inbox census
The census records recruitment source, field dates, question wording, duplicate/exclusion rules and sample size. Results are reported as descriptive findings from that audience unless probability sampling justifies a broader inference.
Privacy and minimum cohorts
Product telemetry is reported only in aggregate. Flick does not publish message bodies, subjects, recipients or individual sender behavior derived from a user mailbox. Small cohorts are suppressed when reporting them could expose a person or customer.
Versioning and raw material
Every released dataset receives a version, collection window, data dictionary and change log. Corrections create a new version rather than silently rewriting the historical file. Download links and licensing appear on the study page when the first measured edition ships.
Current research status
The Exit Gap collection apparatus and sender work list are live, but most senders remain ungraded until the dedicated burner-mailbox observation cycle runs. The swipe benchmark page remains a protocol until a defensible public sample is complete. No missing value is estimated to make a chart look finished.