When it's more important to prevent typos than to cover 100% of all possible email addresses, what syntactic email validations do you do?
You're with the local gas & electric company. 10% of new installations are cancelled because the customer is not home and it costs your company $50 a pop. You start taking email addresses with the new install orders so that you can send reminder emails the day before the appointment. You bring that number to 0% when you have a deliverable email. The only problem is that 15% of all the emails you take bounce back because of some typo. Email addresses are recorded by call center agents who talk to the customer over the phone.
In this case, it's important to collect a valid email address. It's also not all that bad if someone cannot successfully use an email address like:
"Big Momma D.'s #1 CA$H Maker, BOY!!!"@naturalhistory.museum
My guess is that Big Momma D. can't use this email anywhere and either doesn't really use email, or has more conventional email address to use.
This is related to this more general question: How far should I take e-mail address validation?
Let's throw a few more assumptions on here:
- This is in the US, so it's unlike people will have email addresses like 漢字@gmail.com (good luck communicating that to a call center agent in Austin, Texas)
- Validation emails are not suitable, since the customer is only in contact with us by phone