How do I import contacts from one e-mail address to another?

Has anyone published deliverability best practices for mailing to email addresses imported from a users contact address book?

• There are plenty of vendors, ISP-provided APIs and methods for address book importing to use for viral social invites but no one seems to be discussing what happens next after the email addresses are imported to your database. You can't trust the user to be selective in choosing the appropriate contacts and none of the importers seem to have any level of filtering. They import EVERYTHING and if the user selects all then EVERYTHING goes into your mailstream including spamtraps, role accounts, support accounts, sms text addresses etc. I have my own set of obvious accounts to filter out but I'm looking to start a dialogue on better ways to control this.

• Answer:

  I would definitely be interested in your ideas about how this could be regulated. I love hearing fresh perspectives in the delivery world! I'm a postmaster for a mid sized ESP, we allow customers to import contacts via uploading a .csv list, manual entry, or via API. At scale each import cannot be reviewed by a human being so we have automated spam filters that go through and check imports for any indications of spam. Then, if a list sets of a certain amount of triggers we have someone from our abuse/delivery team review the import. If it still looks shady, we hten contact the customer in question and review a few things with them like list source, opt in page examples, list age, frequency of mailing, etc. If it seems all is in order the list is then imported, and we keep a close eye on that accounts delivery for the first few weeks to ensure it is in fact a legitimate list. As for the role account issue, it's very easy to setup filters against them (noreply, postmaster, support) though there is a debate going on about the validity of screening role addresses. The basis of screening these imports is obviously data, so the longer an ESP has been in business it is safe to assume the more spam they have seen and can build a large comparison database of known spam traps or just plain dirty marketers/lists.