New Bounce Handling for Improved Deliverability
In this article
Hard bounce. Anti-spam filters. Delivery rate. Deliverability.
All those who work in email marketing who have never had to face these issues evoking something mysterious and decidedly hostile (starting right from the name), raise their hands.
Those just starting out with the email channel (and they’re not alone) often work with the constant anxiety of the perennially present sword of Damocles – the dramatic chance of winding up in a spam folder. Or, even worse, not even reaching the recipient’s mailbox.
On one hand, this fear is legitimate: the effectiveness of a marketing channel is measured in return on investment, and this is irrevocably undermined by any message that does not reach the desired destination. On the other hand, it is also known that fear often goes hand in hand with lack of knowledge: as human beings we are afraid of what we do not understand, that which we can’t read and interpret, and therefore control.
In the universe of email marketing, deliverability is that area which, more than any other, tops the charts for obscurity and distrust. With this post, I’d like to satisfy two objectives today:
- Clarify once and for all what deliverability truly is and why those who understand it are not afraid of it.
- Demonstrate that optimal deliverability is not a goal, but a starting point (also and above all thanks to the tools and expertise made available by MailUp).
First of all, who am I?
My name is Alberto Miscia and I have been the Head of MailUp’s Deliverability & Compliance department for several years now. And that’s not all: I am one of the company’s five co-founders, which I helped transform from a small company in Cremona into an international group present in over 50 countries.
My work takes place away from the spotlight, behind the scenes, or under the hood as developers and analysts like to say.
The Deliverability & Compliance team creates a technological infrastructure that optimizes message deliverability, minimizes problems with ISPs, and ensures exactly what our customers do not see – the absence of problems, returned messages, and critical issues in mailings. Given that our customers send a hundred million messages a day, it is easy to understand how this goal is particularly ambitious!
Deliverability: what it is and why it scares so many people
The termdeliverability is often confused with delivery (or delivery rate), which is the delivery rate of a message in thestrictest sense. Yet there is a clear distinction between the two:
- The delivery rate indicates how many messages actually reach the recipient’s inbox out of the total number of messages sent. It is a pure and simple metric, easily measurable, whose opposite can be found in the bounce analysis, i.e. the emails rejected by the ISPs.
- Deliverability, on the other hand, indicates a message’s ability not only to reach the recipient’s mailbox, but to avoid the spam folder. It has a much more nuanced measurability which depends on the interaction of a large number of factors.
The most important ISPs have developed and perfected their algorithms over time to determine if the emails received are relevant (inbox) or not (spam). The elements that contribute to this calculation are reputation indicators of the domains used or of certain “fingerprints” of the messages. Here are just a few examples:
- Message sender and subject
- Message structure and content
- Reputation of the sending domain
- Message rate
- Amount of returned messages
- Reactivity of contacts in the database
Why is deliverability so scary? Because it is difficult to understand and easy to damage – all it takes is a few days and a few wrong moves. Improving it is instead a long and delicate process that requires specific skills and ad hoc resources (human and economic). That’s why we work on a double front every day at MailUp:
- Constantly improve the deliverability standard in order to offer all customers a very good basic service.
- Support companies in particularly delicate situations or in the event of critical situations with the support of a specialized team, with the aim of identifying areas for improvement and implementing strategies designed according to the specificities of each case.
Once thesefoundations are clarified, the fear of deliverability disappears:
- Leveraging an already optimized infrastructure such as MailUp removes this concern from the very start: you can confidently rely on a system that guarantees near 99% deliverability in normal situations.
- Mastering best practices lets you avoid making mistakes and actively govern campaign performance.
- And we are here to help for all the rest: if any critical issues arise or there is room for improvement, you’ll have a dedicated team available to support you at every step.
From deliverability to the engagement funnel
When we optimize the deliverability of a message, we lay the bricks of the first, fundamental step of the interaction cycle of each recipient with the email message. We can diagram this cycle into a funnel (similar to what happens in the classic sales “funnel”) and informally call it the engagement funnel.
The challenge faced by those working in email marketing is to minimize losses between one step of the funnel and the next, with the aim of bringing as many users as possible to the crucial action – the click on the message to (finally) land on the desired page.
It is now clear that deliverability is very far from being a sealed compartment or an activity enclosed within exclusively technical logic. Optimizing a message’s delivery is a prerequisite for the subsequent phases of the funnel in order to perform as best as possible.
(Starting from the basics: bounces
In this initial phase of the funnel, the goal is to keep the number of emails delivered as close as possible to the number of emails sent. A bounce is nothing more than an email that is refused directly by the destination server, which then causes a reduction in the number of messages that will reach the second step of the funnel.
Normally, the bounce is received along with a classification code that helps the sender understand its nature (inbox full, non-existent address, etc.).
The bounce is an alarm signal
The immediate question to arise is: why don’t ISPs simply reject the messages, instead of sending them back and explaining the reason for the rejection? The answer is quite simple: to allow senders to act accordingly.
The logical basis here is that informing the sender once will prevent such bounces from recurring in the future, helping to save resources for future occurrences. A sort of courtesy “alert”: if no corrective action is taken following this notification, the ISP will switch from a temporary stop to a complete delivery block, with all the unfortunate relative consequences.
What is thecorrect behavior when emails are returned? The sender should analyze eachsingle bounce to understand whether to include the contact causing the problemin their next campaign.
Warning: a minimum amount of returned emails is normal in any email campaign, and must be managed in an organic and continuous way. This is the first, unavoidable step to solve in order to face the real challenges of deliverability. If, on the other hand, you still manage bounces as an emergency, the road to an articulated approach to deliverability is still quite long.
And if we were to manage all your bounces?
We know all too well that those working in marketing have ambitious goals, little time, and lots of activities to complete. How many marketers regularly carry out the precise task of checking and managing returned emails? Not many (understandably), given the burden and complexity of the matter.
To improve this crucial step of the funnel, therefore, the sending platform itself must see to managing these bounces. Not only in terms of classification and reporting, but at a much more proactive level also in terms of problem solving.
Precisely for this reason, for some time now MailUp has introduced automatic algorithms able to identify the email addresses that generate delivery problems. And that’s not all: these addresses are automatically unsubscribed from the sending database, so as to prevent malicious blocks for the rest of the campaign.
But this was not nearly enough for us. In the last few months we have worked hard to improve the algorithms responsible for managing bounces, adding many more managed cases. A real temporary unsubscription system lets our customers get close to 100% deliverability, without having to risk prematurely losing recipients permanently.
Compared to the past, today our latest-generation algorithms take a greater number of variables into account to calculate when to move a recipient to temporary unsubscription, where the recipient will remain for a period of time related to the type of bounce and the number of times the bounce events occur.
Thisintelligent and flexible calculation makes it possible to adapt bouncemanagement to each individual case, ensuring that contact lists are virtually always clean and effective.
Managing bounces in an intelligent way: a practical example
A few months ago a well-known Italian service provider reported problems that prevented the delivery of all incoming messages with acceptable times. The problem was not only detected by MailUp, but by the entire Email Service Provider community.
Indeed, thisissue greatly limited the number of messages delivered and forced us togenerate the bounce after a few days, as per our policy. Here is what thedelivery and bounce trends in the period in question looked like for a user whowas not yet included in the rollout of the new algorithms:
The number of bounces generated (in blue) was high on the most problematic days, and continued to present itself in the following days. As a result, the open rate dropped significantly, and then returned to slightly lower values than the previous ones.
What happened to those who could insteadalready make use of the new algorithms?
The advanced bounce management allowed us to significantly reduce the sending volumes during the most problematic days, thereby reducing the total bounces and allowing a gradual recovery in the following days.
The most clear difference, however, can be seen in the open rate trend – further proof of how much the different steps of the funnel are strongly interdependent of each other. In the days following the issue, the open rate was even higher than the average of the previous period.
What lies behind the difference in performance? Thanks to the intelligent suspension by MailUp, the provider received fewer messages, users kept their inboxes “cleaner,” and the issue could be more effectively managed.
Intelligent and advanced bounce management improves the performance of email campaigns in the very first phase following sending, laying a solid foundation for optimizing results in the subsequent steps of the funnel.
Contributing to the well-being and better functioning of the email ecosystem has always been part of our ethics. Everyone benefits from better bounce management, in terms of both performance and results (sender, ESP and destination provider side) and in creating a positive, clean, and efficient experience for the end user.