Bounce and block classifications help you make sense of the millions of different SMTP responses that Twilio SendGrid receives from mailbox providers on your behalf. Mailbox providers communicate failures in different ways, and there are many reasons a message may fail. Twilio SendGrid conveniently buckets all of these responses into seven classifications by mapping each unique response to one of them:
This means that you can stop trying to interpret cryptic response codes and messages from mailbox providers and spend your time fixing the root causes of the message failures.
You'll find definitions for each of the classifications and some useful tips and action items for reducing each type of rejection in the following sections.
Mailbox providers modify response codes and rejection messages over time. Our team and our technology work hard to keep these classifications as accurate as possible, but classifications will change as we make adjustments to our system and as mailbox providers change the way they use specific response codes.
The Invalid Address classification groups SMTP response codes that indicate the recipient's email address is syntactically incorrect or the recipient's mailbox does not exist. Email addresses that were valid at some time in the past can become invalid at any time. Some mailbox providers will purge mailboxes that have not been used in a long time. This means that an address could become invalid today even if you were successfully delivering mail to that address yesterday.
Invalid Address rejections are often referred to as hard bounces. Excessive hard bounces can negatively impact your reputation and cause deliverability problems. There is no exact number or percentage of hard bounces that causes deliverability issues, but Twilio SendGrid recommends generating hard bounces from fewer than 5% of your attempted messages.
If Invalid Address rejections make up a large percentage of your total rejections, you can implement several strategies to reduce these rejections. Before taking any action, determine whether or not these rejections recently increased or decreased. Figuring out when Invalid Address rejections became a problem can help you determine the best course of action.
Implement an engagement-based sunsetting policy. Sunsetting policies will help increase engagement rates and help reduce Invalid Address rejections by removing long-term unengaged recipients before their mailboxes become invalid.
The Technical classification is used to group rejections that were caused by a technical problem. Rejections with this classification can indicate a technical problem with the sender or with the mailbox provider. Unfortunately, in most cases, it is difficult or impossible to determine which side the problem is on from the rejection message alone. Here are some common causes for Technical rejections.
Ensure that you have all the necessary DNS records in place. This includes DKIM, SPF, MX, and rDNS records.
Ensure you are passing SPF, DKIM, and DMARC.
Identify which mailbox provider is generating the bulk of these failures.
Technical rejections are more common at smaller mailbox providers than at major providers like Gmail or Hotmail. If most of your mail is flowing to mailbox providers outside of the Big 4 (Gmail, Hotmail, Yahoo/AOL, or iCloud), you should expect higher levels of Technical rejections than someone sending primarily to the Big 4.
Rejections classified as Content indicate that the mailbox provider believes the content of your messages is potentially spam, malicious, or otherwise untrustworthy. Mailbox providers can reject messages for a number of content-related reasons. Here are some common ones.
Suspicious links or URLs within the content.
Message content is too large.
Attachment too large, no attachments allowed, or invalid attachment type.
Mailbox providers can reject messages that contain any attachments. They can also reject messages that contain attachments that are too large or messages that contain specific types of attachments.
Text to Image ratio
Spam or malicious content.
Check your messages for link shorteners. If you need to use a link shortener, make sure it is a custom link shortener that is only used by your organization. Link shorteners are not a problem in and of themselves. However, they are commonly abused by bad actors, so avoid generic or "shared" link shorteners like bit.ly.
Make sure all the links and URLs in your messages are branded. Otherwise, make sure you can vouch for the legitimacy of every domain, link, and URL in your messages.
Check your message sizes. If your messages regularly exceed 5MB, reduce the size of your messages and see if Content-related rejections drop.
Remove attachments from your messages. Attachments do not need to be avoided in all circumstances, but they can cause problems. Send some messages without attachments and see what impact it has on your Content rejections. You can also try sending with smaller attachments or attachments of a different file type.
Ensure your messages have a healthy text to image ratio. There is no firm rule here, but a widely accepted best practice is to include at least two sentences of text for every image in your message. Think of it this way: if your email had no images in it whatsoever, would a recipient know what the message was about? If not, you might have a poor text to image ratio.
If you've investigated the possibilities above, your message content could be flagged because it looks like messages sent by known bad actors. In this case, there is no way to determine exactly what is causing the mailbox provider to believe your message is spam. In almost all cases, you should work on improving how recipients are engaging with your mail rather than attempting to change your subject line or reword your message text. It's much more common for content to be flagged as spam due to high spam complaint rates and poor engagement than it is for content to be "mistaken" for spam because it looks like another message. Fiddling with subject lines and new templates is almost never the answer.
The Reputation classification is used to group rejections that indicate the mailbox provider detects reputation issues with your sending domain or sending IP. Mailbox providers establish your sender reputation using complex algorithms that use hundreds or even thousands of signals to determine your trustworthiness as a sender. In most cases, it's difficult to determine the exact reason a mailbox provider has decided you have a poor reputation. However, your reputation is largely determined by two things: how your recipients engage with your messages and your sending behavior.
Nearly every sender will have some of their mail rejected for reputation reasons. However, if reputation rejections make up a large percentage of your total rejections or you see them on the rise, take action as quickly as possible. Here is a quick checklist for anyone looking to reduce their Reputation-based rejections.
Check your spam complaint rates.
Nothing hurts your reputation more than a user marking a message as spam. If you have spam report rates above 0.1% at any single mailbox provider, take a close look at why recipients might be marking your mail as spam.
Tighten sunsetting windows.
Make sure your hard bounce rates are as low as possible.
Ensure your IP addresses were properly warmed up.
Check your engagement rates.
Consider reducing your sending frequency.
The Frequency/Volume Too High classification groups rejections that indicate the mailbox provider is unable or unwilling to process the number of messages you are trying to send. You are either sending too many messages or sending them too quickly. Every mailbox provider has different capabilities for receiving email. Some, like Gmail, can receive mail about as fast as anyone can send it. However, many smaller mailbox providers lack the bandwidth to accept massive amounts of email very quickly. In these cases, a mailbox provider can temporarily reject email to protect their systems. Mailbox providers can also permanently reject messages if the sender is greatly exceeding the provider's limits.
Some rejections related to sending frequency or sending volume can also fall under the Reputation classification. Sending volume and frequency limits are often tied to reputation. Twilio SendGrid's systems do a great job of separating reputation-based rejections from non-reputation-based rejections, but the difference is sometimes very difficult to accurately determine.
If a high percentage of your total rejections are in the Frequency/Volume Too High classification, determine which mailbox provider is returning these rejections.
Once you have identified the affected mailbox provider, you should reduce the number of messages you are sending to that mailbox provider on an hourly, daily, or weekly basis. This could mean trimming some unengaged recipients from your mailing list or eliminating a few of your daily sends to that mailbox provider.
If you are unable to reduce the amount of messages you send to that mailbox provider, you should determine if you can spread your messages out over a longer duration. Instead of sending one large campaign at 1:00pm, you can spread that campaign over four smaller sends throughout the day.
If these actions do not correct the problem, review the How to address Reputation rejections section of this page. As mentioned before, there are circumstances where your reputation causes a mailbox provider to cap how much of your mail they will accept. We do a good job of properly classifying reputation-based rejections like this, but classifying rejections is an inexact science. Additionally, it's never a bad idea to take some small steps that will help improve your reputation.
Some rejection messages are indecipherable or ambiguous. Most mailbox providers return helpful, well-written rejection messages that give you some idea of what the problem is. Some mailbox providers do not. Rejection messages like "550 There was a problem" are common. When we receive rejection messages that are this ambiguous, we place them into the Unclassified bucket.
The best way to deal with excessive Unclassified classifications is to use the Email Activity APIs or the Event Webhook to analyze any bounces or blocks that were marked as Unclassified. Sometimes there are clues or insights in the response string that our system wasn't able to pick up. We are constantly working on more accurate classifications. As our system learns, the number of "Unclassified" rejection messages gets smaller every day and will continue to improve over time.