Apr 18, 2026 5 min read
The case for slower software
Not slower to load — slower to demand. Software that moves at the speed of a person thinking.
Let me be precise, because "slow software" sounds like a bug report: I want software that loads instantly and then waits. The speed problem in most products isn't latency — it's urgency. Everything arrives fast and then immediately wants something back.
Urgency is a design material
Autoplay, streaks, unread counts, expiring offers — these are urgency mechanics, and they work, in the way that shouting works. Products use them because attention captured today shows up in this quarter's chart, and trust lost slowly shows up in nobody's.
But watch someone use a product built on urgency. They move fast and shallow, clearing states rather than doing work. The interface has set the tempo, and the tempo is panic.
Setting a human tempo
The alternative isn't friction — it's patience. A weekly rhythm instead of a daily streak. Seven good recommendations instead of an infinite feed. A digest that arrives when you asked, not when an algorithm got nervous about your engagement.
When we rebuilt a music app around exactly this — seven albums a week, no autoplay, nothing expiring — listeners didn't leave. They finished more of what they started. It turns out people can tell when software respects their pace, the same way you can tell when a conversation partner is actually listening.
Fast to respond, slow to demand. That's the tempo I'm designing for now.