A few months ago I found out about a Salesforce default change from Winter '21. (Yes, I guess I was late to the party learning about a change from Winter '21. But I've seen several posts from others just starting to figure things out. So...)
Starting with that update, users in new orgs will have their default email encoding set to UTF-8, otherwise known as "unicode.". Prior to that the default was ISO-8859-1 (aka Latin1). That's good, because unicode is a much broader standard, with support for almost all possible characters. (Plus it includes emoji—and we know I love me some emoji! 😘)
If you ever wondered why an outbound email from Salesforce did something strange, such as drop emoji that you know you put in the title, this is probably the culprit. (Yes, that's exactly how I started digging into this.)
Check Yourself Before You Wreck Yourself
Check your own user setting to see if you're still in ISO-8859 (Personal Settings > My Personal Info > Language & Timezone > Email Encoding) and, if you are, Step 1 is to set that to Unicode (UTF-8). You won't regret this change, I promise.
Now have fun sending an email replete with emoji! 🏴☠️
Future Proof Your Org
Admins reading this blog should probably consider fixing the problem for everyone. Go to Setup>Users and check the email encoding of all your users. If your org is more than a couple of years old, you probably have a lot of users that are still on Latin1. You can—and I would argue should—just switch them all in bulk. I really don't think there is any downside. UTF-8 is backwards-compatible with ASCII and nearly any email client being used today will have no trouble with it.
Options for updating the users in bulk depend on what tools you have available. If you've got Apsona it will take you longer to read this article than to make the fix. You can also do this via Dataloader.io or other tools.
Now For The Cool Part
I haven't been able to verify this with documentation, but I think that you only have to make this change and not worry about it going forward. As far as I can tell, once you've set users to UTF-8, when you go to create a new user, they are going to default to UTF-8.
(Or maybe it's that all new users already started defaulting to UTF-8 with Winter '21 and I just didn't notice that until after I had set existing users to UTF-8? But I could swear I had newer users created in client orgs that were still on Latin1. I even had set myself a task to build a flow for new users. But before I built it I started noticing that new users were defaulting to unicode... 🤷🏻)
But the bottom line is that you don't have to remember and you don't have to build a flow to keep things in line. So hooray for that! But make sure you update all those existing users.