Working notes on the decisions behind the product — mostly the ones where the cautious option was also the unpopular one.
Nine months of data on how often the ladder reaches its third rung, and why the number going down is the metric we actually optimise.
Notifying everyone at once feels safer and performs worse. What we learned from the accounts that reached +36h, and why rank order matters more than roster size.
The fourth time we were asked to shorten the vault delay, and the reasoning we now paste into every one of those conversations.
Grace windows, travel holds, and the difference between a system that punishes inattention and one that survives it.
Every support ticket we cannot resolve, listed honestly, alongside why we keep the constraint anyway.
There isn't one. Entries are posted here and nowhere else, roughly monthly, and the RSS feed is the only way to follow along — which suits a product whose entire premise is not bothering people unnecessarily.