Yesterday we deployed a tiny feature. In some obscure pull down menu of our application we've added an option to the users for downloading a more detailed report. We've implemented this feature on a request of one user who requested such report and we thought also other users could benefit from it so we released this feature for all the users. But considering its obscure placement and nature, we didn't think it necessary to release the feature gradually to beta users first as we normally do, but we released it to everybody. What a mistake! One of our more important customers found out the feature later in the day but when she wanted to generate the report, our application miserably failed with a 504 gateway timeout error.
I find it time and again that there are no shortcuts in product development. If you're not willing to invest sufficient time and energy in some feature, you are better not developing it at all. It will be much better for your users who will not be relying on broken tools and it will be much better for yourself since you won't be spending your days extinguishing fires caused by poorly designed and/or implemented features. And yesterday I've learned that regardless how obscure and little a feature seems to you, there's no such thing as a tiny feature!