It has happened often enough that you have stopped noticing. You open a pull request with four hundred lines in it, and within two minutes a colleague has approved it. No comments, no questions, just a green tick. It feels pleasant at first. Then one day something you were unsure about goes to production, breaks, and you realise nobody ever really looked.
An approval that costs nothing tells you nothing. It is not quite the reviewer's fault. They may be busy, they may trust you, or they may feel that questioning a more experienced colleague is not their place. Whatever the reason, you are not getting the feedback you need, and that is something you can change.
Start with what you send. A large pull request with no description invites a quick approval, because there is nowhere obvious to begin. Make it smaller where you can. Then write a few lines at the top saying what the change does, why, and how you tested it.
Most importantly, say where you want their attention. I am least sure about the retry logic in the second file. Does the error handling here make sense to you? Could this query be slow with a large customer? A specific question is much easier to answer than a general request to review, and it tells the reviewer that you actually want them to find something.
If the quick approvals continue, talk to the person directly, privately and without blame. Something like, I really value your reviews, and I would find it useful if you pushed back more. I am sure I am missing things. Many people approve quickly because they assume you want speed. Telling them otherwise is often all it takes.
Thank people when they do find something. If the first time someone raises a concern you argue with them, they will go back to the green tick.
And look at your own reviews. If you approve things quickly too, the team may have quietly agreed that review is a formality. Changing that starts with somebody asking for more, and it may as well be you.
– Asael Shinder
Top comments (0)