What kind of mail do filters target?

All to often we think of filters as a linear scale. There’s blocking on one end, and there’s an inbox on the other. Every email falls somewhere on that line.
Makes sense, right? Bad mail is blocked, good mail goes to the inbox. The bulk folder exists for mail that’s not bad enough to block, but isn’t good enough to go to the inbox.
Once we get to that model, we can think of filters as just different tolerances for what is bad and good. Using the same model, we can see aggressive filters block more mail and send more mail to bulk, while letting less into the inbox. There are also permissive filters that block very little mail and send most mail to the inbox.
That’s a somewhat useful model, but it doesn’t really capture the full complexity of filters. There isn’t just good mail and bad mail. Mail isn’t simply solicited or unsolicited. Filters take into account any number of factors before deciding what to do with mail.

What kinds of factors?

There are five broad questions I think about when guiding clients through their email programs.

Is the mail safe?
Is the mail solicited?
Is the mail targeted?
Is the mail wanted?
Is the mail productive?

Different filters have different weights for the categories. Those weights explain why delivery can range so widely across domains and email providers.
Let’s look at each set of factors and talk about who might care more about those factors than others.

Is it safe?

Does the message contain malware, phishing, anything that could harm the recipient’s computer or the network as a whole? These filters are widespread and heavily weighted by most people. Safe doesn’t typically come into it for legitimate mail, but the filters are still there and still sniff at our mail.

Is it solicited?

Alternatively, did the user ask to receive mail from the sender? Many blocklists, including Spamhaus, specifically set out to block unsolicited email. They don’t really care about what the email is. They simply want to make sure that the recipients are receiving mail they asked for.
Confirmed opt-in is a way to ensure that mail is solicited. The folks behind many of the blocklists simply want users to receive mail they asked for. Senders who can demonstrate the mail is solicited get removed from the list.
At ISPs, solicited is somewhat important, but the signs of solicited mail overlap with signs of wanted mail. When ISPs measure unknown users and complaints, part of what they’re trying to determine is if the mail is solicited by their user.

Is it targeted?

Does the user understand why they’re receiving the mail? As a small business owner, I get a lot of targeted email. Random companies buy addresses and target me as someone who might want their service. The mail is targeted, so some filters, particularly those at ISPs, might not block or spam folder the mail.
But just because mail is targeted doesn’t mean the user wants it.

Is it wanted?

Does the user want the mail? Sometimes they do, sometimes they don’t. The big webmail providers (Oath, Microsoft, Gmail) heavily weight wanted. They don’t care so much if the message is solicited or targeted, although both things will increase the likelihood that the mail is wanted. At these ISPs, filters really focus on signs that the user is engaged with the message as part of the delivery process. Wanted mail gets into the inbox, unwanted mail not so much.
But just because the mail is wanted doesn’t mean it will make it to the inbox.

Is it productive?

This filter only really comes into effect when we’re talking about mailing into businesses. Email is a tool for businesses and they often want employees to be working while at work. Even if an employee solicits and email a business might decide it’s not productive for the business and they block that source of email. Likewise, businesses will block targeted and wanted messages simply because they’re unproductive.

What’s it all mean?

Effectively addressing delivery problems means understanding why a message isn’t reaching the inbox. Improving engagement isn’t going to help senders reach employee mailboxes if the mail is unproductive. Better targeting won’t help if the block is due to the mail being unsolicited. Using confirmed opt-in won’t magically get malware into the inbox.
It used to be that deliverability recommendations would work across the range of filters. Mail that made it to the inbox at an ISP like Gmail was likely to make it into the inbox almost anywhere. But as Gmail (and Oath and Microsoft) focus more and more on custom delivery for every recipient, recommendations that work there aren’t always going to work elsewhere.
Reaching the inbox outside of webmail providers means taking a lot more into account than just if the recipient is engaged with your mail.

July 2015: The Month in Email

Once again, we reviewed some of the ways brands are trying (or might try) to improve engagement with customers. LinkedIn, who frequently top lists of unwanted-but-legitimate email, announced that they’ll be sending less mail. Josh wrote about giving subscribers options for both the type and frequency of messages, and about setting expectations for new subscribers. In each case, it’s about respecting that customers really want to engage with brands in the email channel, but don’t want the permission they’ve granted to be abused. I also wrote a brief post following up on our June discussion on purchased lists, and as you’d predict, I continue to discourage companies from mailing to these recipients.

Email predictions for 2015

Welcome to a whole new year. It seems the changing of the year brings out people predicting what they think will happen in the coming year. It’s something I’ve indulged in a couple times over my years of blogging, but email is a generally stable technology and it’s kind of boring to predict a new interface or a minor tweak to filters. Of course, many bloggers will go way out on a limb and predict the death of email, but I think that’s been way over done.

Even major technical advancements, like authentication protocols and the rise of IPv6, are not usually sudden. They’re discussed and refined through the IETF process. While some of these changes may seem “all of a sudden” to some end users, they’re usually the result of years of work from dedicated volunteers. The internet really doesn’t do flag days.
One major change in 2014, that had significant implications for email as a whole, was a free mail provider abruptly publishing a DMARC p=reject policy. This caused a lot of issues for some small business senders and for many individual users. Mailing list maintainers are still dealing with some of the fallout, and there are ongoing discussions about how best to mitigate the problems DMARC causes non-commercial email.
Still, DMARC as a protocol has been in development for a few years. A number of large brands and commercial organizations were publishing p=reject policies. The big mail providers were implementing DMARC checking, and rejection, on their inbound mail. In fact, this rollout is one of the reasons that the publishing of p=reject was a problem. With the flip of a switch, mail that was once deliverable became undeliverable.
Looking back through any of the 2014 predictions, I don’t think anyone predicted that two major mailbox providers would implement p=reject policies, causing widespread delivery failures across the Internet. I certainly wouldn’t have predicted it, all of my discussions with people about DMARC centered around business using DMARC to protect their brand. No one mentioned ISPs using it to force their customers away from 3rd party services and discussion lists.
I think the only constant in the world of email is change, and most of the time that change isn’t that massive or sudden, 2014 and the DMARC upheaval notwithstanding.
But, still, I have some thoughts on what might happen in the coming year. Mostly more of the same as we’ve seen over the last few years. But there are a couple areas I think we’ll see some progress made.

April 2016: The Month in Email

We are finishing up another busy month at WttW. April was a little nutty with network glitches, server crashes, cat woes, and other disruptions, but hopefully that’s all behind us as we head into May. I’ll be very busy in May as well, speaking at Salesforce Connections in Atlanta and the Email Innovation Summit in Las Vegas. Please come say hello if you’re attending either of these great events.

Speaking of great events, I participated in two panels at EEC16 last month. We had a lot of great audience participation, and I met many wonderful colleagues. I wrote up some more thoughts about the conference here. I also had a nice conversation with the folks over at Podbox, and they’ve posted my interview on their site.
In the Podbox interview, as always, I talked about sending mail people want to receive. It always makes me roll my eyes a bit when I see articles with titles like “5 Simple Ways to Reach the Inbox”, so I wrote a bit about that here. In addition to sending mail people want to receive, senders need to make sure they are collecting addresses and building lists in thoughtful and sustainable ways. For more on this topic, check out my post on list brokers and purchased lists.
These same not-so-simple tricks came up again in my discussion of Gmail filters. Everyone wants a magic formula to reach the inbox, and — sorry to burst your bubble — there isn’t ever going to be one. And this is for a good reason: a healthy filter ecosystem helps protect all of us from malicious senders and criminal activity. The email channel is particularly vulnerable to fraud and theft. The constant evolution of filters is one way mail providers can help protect both senders and recipients — but it can be challenging for senders and systems administrators to keep up with this constant evolution. For example, companies sometimes even inadvertently filter their own mail!
I also wrote a bit about how B2B spam is different from B2C spam, and how marketers can better comply with CAN SPAM guidelines in order to reach the inbox. We also republished our much-missed friend and colleague J.D. Falk’s DKIM Primer, which is extremely useful information that was at a no-longer-active link.
One of my favorite posts this month was about “dueling data”, and how to interpret seemingly different findings around email engagement. We also got some good questions for my “Ask Laura” column, where we cover general topics on email delivery. This month we looked at “no auth/no entry” and the Microsoft Smartscreen filter, both of which are useful things to understand for optimizing delivery.
Finally, we are pleased to announce that we’ve joined the i2Coalition, an organization of internet infrastructure providers. They posted a nice introduction on their blog, and we look forward to working with them to help advocate and protect these important technical infrastructures.