A year is long enough to lie to yourself about ("I've got months left"); a week is too short to fit anything ambitious into. Ninety days sits in the gap that actually works: long enough for a real objective to take shape across a few phases, short enough that you can still feel the deadline by week six.
This isn't about a framework — that's the OKR page, if what you're after is objectives and key results — it's about the timeframe itself. The manual builder maps a 90-day goal onto phases and milestones, lays them out on a timeline so you can see the quarter at a glance, and runs the weekly cadence underneath so each week has its own actions rather than the goal sitting untouched until week 12.
Ninety days is also short enough that a bad two weeks doesn't sink the whole thing (there's still time to adjust before the quarter's out), which is part of why it works better for a lot of people than an open-ended year-long goal. Just try it: pick something you can actually finish in a quarter, and let the timeline show you what each week needs to look like.