Engagement drives deliverability

Return Path released an white paper today offering the Secrets of Successful Senders. I don’t think any of my readers will be surprised that it boils down to identity, reputation, and engagement. Return Path treats these as separate things and I understand why they do. I think however, that the identity and reputation are supporting players to the overarching issue of engagement.

When I’m dealing with clients and troubleshooting deliverability problems and offering solutions, I focus on the root cause. To me the root cause is almost always a data problem. Either there’s a problem with data collection or there’s a problem with data maintenance. These problems result in mail going to people who don’t really want or care about it.
Yes, identity is important. But, realistically, anyone mailing through a decent ESP has SPF and DKIM in place, at least on some level. There may be better ways to authenticate, but the boxes are checked.
Yes, reputation is important. But here’s the thing, reputation just means that the ISP knows how users are going to react to an email. Reputation isn’t some nebulous concept made up by ISPs. It’s an actual measurement. It quantifies the history of an IP or a domain or a mail stream and says we know that this IP sends wanted mail. We know that this domain sends mail our users ignore. It’s a history. Past performance does indicate future results.
Identity says who a sender is. Reputation tells us that sender’s history of sending. Those are the two factors that enable ISPs to make delivery decisions. Mail comes in and the ISP looks at it. They use identity to determine what reputation to assign to a mail. Reputation drives delivery, whether into the inbox or the bulk folder.
 

Engagement is the key

All of the various metrics we use are proxies for engagement. High levels of complaints tell us users don’t like the mail. Low levels of opens tells us they don’t care enough to read it. Excessive bounce rates tell us that our data collection process is poor.
I see too many deliverability consultants and experts suggest that the fix is to make the numbers look better. Lower overall bounces using a list hygiene company. Remove complainers by using multiple ESPs. Use gimmicky subject lines to get recipients to open. These processes will sometimes fix delivery in the short term. They’re a solution in specific cases, and typically work for companies with high brand recognition and a core of engaged users.
For companies without a core of engaged users, however, all they end up with is a list that has low bounces, low complaints and erratic deliverability. They don’t know how to address the problem because all they’ve done is hide the symptoms.
If I take aspirin for a fever, I haven’t cured the flu. I have just lowered my fever. If I use a list cleaning service to remove bounces, I haven’t suddenly found a list of engaged users. I’ve just lowered my bounce rates.

Why purchased lists don’t work

Lack of engagement is the underlying reason purchased lists don’t work. Companies sending to purchased lists have identity. Sometimes, they even have a decent reputation built on the back of actual opt-in emails. What they don’t have is engagement. The majority of recipients don’t really care about the mail, so they aren’t engaged. Therefore the messages never make it to the inboxes.
The dirty little secret of list sellers is they work very hard to remove negative signals from their product. They’ll remove bouncing addresses, they’ll even try to remove users that will complain about the mail. The measurements say this is a “good” list, but it’s lacking one thing. The recipients on that list don’t necessarily want mail from the final purchaser.

Putting it together

Yes, identity is important. It’s how the filters know what the mail’s reputation is. Good reputation is important, but the only way to build a good reputation is send mail that’s wanted. How do ISPs know what mail is wanted? Their users tell them so.
Engagement is the root of all deliverability. Everything else just makes it easier for filters to see the engagement.
 

Related Posts

March 2017: The Month in Email

It’s that time again… here’s a look at our last month of blog posts. We find it useful to recap each month, both to track trends and issues in email delivery and to provide a handy summary for those who aren’t following along breathlessly every single day. Let us know if you find it useful too!

As always, I wrote about email filters. It’s so important to recognize that filters aren’t arbitrary — they’re detailed instructions that help meet specific user needs, and the more you are cognizant of that, the better you’ll be able to work with them. Additionally, filters aren’t perfect and likely never will be. False positives and false negatives are frustrating, but as long as spam is still a viable business for spammers, they’ll continue to figure out how to work around filters. As such, we can’t expect filters to be 100% accurate in determining what constitutes wanted and unwanted mail.
Part of this, of course, is due to the problem of fraudulent signups. Companies aren’t particularly vigilant about address acquisition and hygiene, and as a result, they’ll claim you “signed up” for their email when you did not. Some people believe that a confirmed opt-in (COI) will solve this problem, but our experience is companies are reluctant to leave revenue on the table, and that they will continue to mail to addresses that have not confirmed.
Address sharing and co-reg is also part of the problem. As we saw in the extensive RCM data breach, many major brands continue to work with third-party senders to send mail in ways that are quite clearly spam. And in more criminal activity, I looked at the rise of botnets and how some of those criminals were brought to justice. In other justice news, there’s been an indictment in the Yahoo breach and another CASL enforcement action.
I wrote a post about bounce handling and “relaying denied” error messages, which are quite rare. It’s useful to have an understanding of these and other error messages, since bounces are sometimes indicative of a larger technical issue, such as when AOL accidentally bounced all messages for a short period last week. Speaking of AOL, we noted that there’s no official timeline for the move from Verizon addresses to AOL addresses following the 2015 acquisition, but it may be worth considering asking your customers to update their addresses.
Spam and filters aren’t the only factors of course. It can be challenging to figure out the multiple factors that make up the black box of delivery. And of course, the most important part of delivery continues to be engagement, engagement, engagement.
I wrote a few posts this month on why I do what I do, and why it’s so important to me. First, I wrote about A Day Without A Woman, and my choice not to participate in offering advice and guidance for that day. The truth is that I enjoy sharing what I know and helping people solve problems. I was honored to be named one of 11 Innovators in Email, and I know that my volunteer work in the industry and my unpaid blogging work is a big part of that. It may sound corny, but I really do believe we are on the front lines of the fight of good vs. evil online, and despite the distractions of politics and world events, we must all continue to do our part.

Read More

You're kidding me

All the authentication and DMARC in the world can’t save you from stupid.
I just got a survey request from my bank. Or, at least, it claimed to be from my bank.

Read More

Phishing protection

Last week Return Path announced a new service: Domain Assurance. This service allows companies who send only authenticated email to protect their brand from phishing attacks. Participating ISPs will reject unauthenticated email from domains participating in this program.

Read More