Ask HN: Async updates: Input friction vs. Output friction?

I'm building a tool for async standups, but I'm torn on where to focus the automation first. I see two distinct failure points:

1. Input Friction (The Maker's Tax)

Context Switching: Stopping coding to recall yesterday's work and writing a paragraph is a huge flow-killer for devs.

Result: Updates become vague ("fixed stuff") or treated as a mindless chore.

2. Output Friction (The Manager's Tax)

Signal-to-Noise: Managers have to parse scattered logs across Slack/Jira to reconstruct the actual project state.

Result: Updates are posted but ignored because consuming them takes too much cognitive effort.

In your experience, which bottleneck is the bigger problem?

Should I prioritize making it easier to write (auto-drafting from logs) or easier to read (auto-summarizing context)?