I had it happen to me. We had an SSL certificate for a website that one of my employers ran. I actually purchased the certificate, and it secured our communications for a couple years until one day it didn't. The certificate expired and the website stopped accepting connections. If I had noticed, it might not have been a big deal. However when the owner of the company gets a call from one of his large customers, it's an much bigger issue.
If you worked at Microsoft last year when their security certificate expired, it was an even bigger deal. It's not that an affected customer that sends a note; it's affected, unaffected, and potential customers that hear about the issue from the media. It seems like tracking private virtual properties (PVPs) ought to be easy, but it's not. As pointed out in this piece, there are a number of issues at an organizational level, and while there are fixes, it takes some effort.
In many businesses that have periodic activities required for physical assets, there are often people dedicated to tracking, or performing the activity. Mechanics know about maintenance on vehicles, accountants renew leases, workers replace equipment nearing the end of service. Often the time lines and activities involved are understood, and individuals understand their responsibilities.
This showcases another area in which technology is woefully immature. Need a certificate? It takes a person with specialized knowledge to understand what's needed, purchase it, and install it. This person either then moves on to a new role without leaving instructions behind, or isn't well equipped to understand the need to track the expiration and replacement of the technical item. There often isn't even a system set up to handle replacements of these items, which might be superseded or replaced by some entirely new type of technical wizardry.
Managing and tracking PVPs is hard, and I suspect, going to get harder. Security requirements increase, technical requirements grow, and specifications change. I'd like to say I'm confident Azure (or AWS or another large service) will never have another outage because of this, but I wouldn't be surprised if they do.