Code Reviews Shouldn’t Take Days
A review that takes three days is rarely thorough. It’s just late.
We’ve all seen it.
A pull request opens on Monday, gets reviewed on Thursday, it gets re-reviewed the next Tuesday (after several rounds of nitpicking and “just one more thing…”).
By then, the original context is gone, the developer is already two tasks ahead, and the change feels risky again.
The Cost of Delay
Code reviews are about communication, trust, and momentum, not just about catching bugs and making sure no console.logs are in the PR.
When we wait too long:
- Context fades: reviewers forget what the code is about.
- Frustration builds: devs feel blocked and undervalued.
- Quality drops: out-of-sync branches cause merge hell.
Quality ≠ Perfectionism
A fast review doesn’t mean a shallow one: it means being clear on what matters:
✅ Architecture
✅ Long-term maintainability
✅ Alignment with existing patters
Not:
❌ Personal syntax preferences
❌ “I wouldn’t have done it like this” remarks
❌ Endless tiny rounds of back-and-forth
Practices That Work
Here’s what high-trust teams do instead:
- Treat reviews as blockers, not background noise.
- Timebox feedback loops: e.g. max two rounds before pairing up.
- Leave clear comments: suggest improvements, not just critiques.
- Push the team norm: fast review is a sign of respect, not pressure.
TL;DR
- Code reviews are not critique but are collaboration.
- Speed matters of course, but so does focus.
- Review today, not next week.
💬 How does your team handle reviews? What helps keep them productive and kind?
In case you want more guidance: download my Free Developer Growth Kit — 3 Practical Guides to Grow from Coder to Leader. It gives you practical tools to level up your skills, mindset, and workflow, without burning out or guessing your next step.
Top comments (2)
Practices i would add are:
I find it weird when people want to learn from code written by people they don't know, but don't want to learn from people they work with.
As a reviewer I'm always curious when I see code I didn't expect from the feature or problem information.
@xwero Thank you for your tips and joining the conversation, much appreciated! I especially like the part about big mistakes or many mistakes!