If you're pushing changes to production many times a day, you can use realtime metrics to notice if anything goes awry, and know when you should roll back. It w

If you're pushing changes to production many times a day, you can use realtime metrics to notice if anything goes awry, and know when you should roll back. It won't give you the same nuances as an A/B test over a month, but it's a great way to do a quick double check to keep things moving quickly while avoiding disaster.