On Prediction Accuracy

I purchased an iPhone 15 Pro Max for my wife on September 23rd and was initially informed that the delivery window would be between October 24th and October 31st. About a week later, I received an email updating the expected delivery to October 23rd to October 30th. Then, after another week had passed, I received another email indicating that the new delivery date would be October 16th. To my surprise, today, on October 11th, the new phone was delivered to my house. I am absolutely thrilled that I received the phone nearly two weeks ahead of the original promise. I have always been a strong advocate of the "under-promise and over-deliver" approach. In fact, I wrote about my admiration for Amazon's consistent practice of under-promising and over-delivering over a decade ago in a blog post.

However, in this case, Apple's initial estimate was significantly too conservative. It's clear that their prediction model's accuracy has significant room for improvement.