Schroedinger’s email

The riskiest email to send is that very first email. It’s a blank slate. Even if you’re sending confirmation messages, you don’t really know anything about how this email is going to affect your reputation.

It’s Schroedinger’s email. The address is both good and bad, until you send to it.

If it’s good, great things will happen. You’ll be happy. The recipient will be happy. Deliverability will increase. Everything is awesome.

But if it’s not good, there are a whole host of consequences to sending that email. The obvious problems are hitting spamtraps, bouncing mail and complaints with the resulting delivery problems and, in very bad case, spamfoldering and blocklisting. Sure, you can use a data hygiene to lower the chances of the mail bouncing. But hygiene services don’t help you if the address is deliverable but belongs to someone else. They certainly don’t help if the address is a spamtrap.

There are all sorts of ways to mitigate damage from bad email addresses, after you know for certain they’re bad email addresses. But that first email is still risky. Even sending a confirmation email (double opt-in) can cause delivery problems at some places. That’s especially true for signups where you might want to send confirmation emails like sweepstakes or political mail.

The obvious answer is to segregate all confirmation emails onto their own IP with their own DKIM signature and, if you’re really worried, it’s own domains and everything. The problem there is that if your mail is messy enough, you may generate a bad reputation on it and your confirmation emails will go to bulk.

You may want to consider, then, just mixing in the confirmation emails with your regular mailstream and letting the good reputation carry the new messages. That may work depending on the relative volumes and the quality of the subscription feed.

Another way to handle it is to segregate the confirmation messages on an IP with other transactional and triggered emails like password resets, 2FA emails and purchase receipts. If you have a transactional feed, this is the best way to handle this mail. Most of the other emails are heavily engaged with, but come at irregular intervals. This mimics the confirmation emails and lets all that stream develop a reputation outside of the reputation of regular bulk mail.

All in all, there’s no one way to manage confirmation emails for a signup stream. There’s always going to be risk to mailing that unknown email. We’re already seeing filters able to sort out different mail types when they’re from the same IPs with the same authentication. Google and Oath are good at that already.

My best advice is to lump it in with the other transactional email. That’s what it is, that’s what it looks like. If you only have a single IP, then I’d advise authenticating transactional mail, including confirmation mail, differently from marketing and bulk mail. That way the filters can distinguish between the two streams. While some reputation will be shared between the different kinds of mail, the filters will be able to distinguish between them. As such the confirmation emails will be less likely to harm your overall delivery.

Related Posts

Brand indicators in email

A number of companies in the email industry have been working on a way to better identify authenticated emails to users. One proposal is Brand Indicators for Message Identification (BIMI). A couple weeks ago, Agari announced a pilot program with some brands and a number of major consumer mail providers. These logos should be available in the Yahoo interface now and will be rolling out at other providers.

Read More

Data Integrity, part 2

Yesterday I blogged about eROIs contention that consumers should not be wasting the time of lead gen companies by filling in fake data. There were lots of good comments on the post, and I strongly encourage you to go read them if you are interested in different perspectives on the data issue.
One of the arguments I was making is that people are only going to give accurate information if they trust the website that is collecting information. I do, strongly, believe this. I also believe very strongly that websites collecting information need to do so defensively. It is the only way you can get good information.
This ties in with an earlier post about a website that collects email addresses from any visitor, then turns around and submits those addresses to webforms. Hundreds of mailing lists have already been corrupted by this group. They are a prime reason companies must design address collection process defensively. There are people who do bad things, who will take an opportunity to harass senders and recipients. This company is not the first, nor will they be the last to commit such abuses.
Taking a stand against abusive companies and people may be useful, but that will not stop the abuse. It is much easier to design process that limits the amount of abuse. For lead gen, in particular, confirmed opt-in is one way to limit the amount of bad data collected. As a side effect, it also results in less blocked mail, fewer complaints and better delivery.

Read More

Confirming addresses for transactional mail

A colleague was asking about confirming transactional mail today. It seems a couple of big retailers got SBLed today for sending receipts to spamtraps. I talked a few weeks ago about why it’s important to let people unsubscribe from transactional email, and many of those same things apply to confirming receipts.

Read More