Mind sharing how you're achieving this? I've been looking for a simple solution for this for a while.
I'll pass along how –I– do this. My hosting provider, Pair.com, uses qmail as a mail server, and qmail supports Sub-Address Aliasing through its qfiltering mechanism. So I can create a "recipe" of the form "'account'
-default@mydomain.tld" that accepts mail addressed to the "account" dash "anything" address. I can then filter locally on the "anything". I create a bunch of rules for TLDs, like "com-default", "org-default", "info-default", etc, and have them dumping the inbound mail into an IMAP folder of my account, while blocking the com@/org@/info@ main account addresses.
For example: a 'com-default' rule would catch
com-amazon@mydomain.com, and I can give that address to Amazon, using their top-level domain, without me having to make a
specific Amazon rule. If/when Amazon loses their email database to spammers (better example might be Zappos, or Dropbox, or Adobe, all who lost their email databases), I can first put heavier anti-spam rules on JUST "
com-amazon@mydomain.com", again using the qmail rules mechanism. Could say, for instance, that if the sender for that specific address isn't "from @amazon.com", then trash it. Or, if the spam gets too bad, I can just block that specific address entirely, without effecting other "com-*" addresses! Finally I block all mail to "
com@mydomain.com".
This gives me a ton of flexibility. I just use the B2C's domain as a "unique" address AND I get to see who is selling my addresses. Or has had a security breach.
Gmail also supports sub-address aliasing, using the "+" delimiter. However, it is terribly ineffective because Google (stupidly) doesn't let you select an alternate "account" (the front part, before the +) for a gmail address, so any spammer that sees
account+something@gmail.com knows that account is a legit Gmail account, can drop the +something, and hit the Inbox directly. You can kinda wrangle around this with creative use of "." in the address, but, again, spammers know that they can just drop "." without consequence. Though it does give you a bit of leverage: if you use, say "
f.roo+amazon@gmail.com" and start getting mail at "froo+amazon@gmail" or just "
froo@gmail.com", you know it is spam (since you can assume if they're dropping "." and "+*" they're up to no good).
Outlook supports sub-address aliasing too, similarly using "+", but again suffers from the same stupidity as Gmail and doesn't, I don't think, allow "." usage. Also, it doesn't work on 365 Business addresses. Nor can you can't address aliasing with M365 Personal/Family.
However many web forms won't even ALLOW "+" in the account portion of an email address, which technically violates the RFC, since + and - are allowable address characters. I have had better luck with "-". IIRC, Comcast does this. (Comcast won't even let you USE an email address with "comcast" in it.)
I had, long ago, mused upon a much better solution: using an encryption algorithm to generate addresses on-demand. When Apple came out with their Sign in with Apple ID email thing, I had high hopes. Unfortunately, their solution falls short since the entire system hinges on mail redirection within SMTP (which is fraught with danger). The best solution is one that would run on the front end of your mail server, using your domain, and be able to quickly reject blocklisted inbound addresses. Although black-holing them is also a valid solution too. And I want encryption so that addresses can't be easily guessed. Apple really is the only one that can do this effectively, since they have the predominant position at the mail client and the browser. Need the mail client to "unwrap" the address portion for local filtering as well as report back "spam" addresses to the blacklist on the server. Need the browser to be able to generate the unique addresses in webforms and push that data to the server/password list. And finally need an open-sourced algo to generate the addresses which could be deployed to mail servers. The best solution would tie together the encrypted address with SPF records; if a message's decrypted address's domain doesn't yield a SPF record matching the message's sender, that's a pretty good indicator it is spam. But, going back to the SMTP rerouting, that can't be known for sure. And it requires a good bit of computation upfront. (But, IMHO, better than wasting my time with spam.)