Comprehensive Guide to Deliverability and Inbox Placement

Last updated: April 2024

Email Deliverability: What is this?

Mailbox providers are not required to deliver your mail, and may delay receipt or reject it outright if they believe that you're sending unsolicited bulk email, also known as spam.

Email deliverability is a general term that refers to whether your emails are being delivered to your supporters. There's a great deal of overlap with the concept of Inbox placement, which is concerned with whether your emails are placed into the inbox instead of spam, promotions, or another folder (see below).

In 2022, ActionKit ran a deliverability training with info that's still highly relevant; you can watch the recording at https://www.youtube.com/watch?v=gVuABY5LW00

Inbox placement: Will your email end up in spam or the inbox?

Mailbox providers determine whether your email ends up in spam or the inbox through a wide variety of factors, but the most important factors boil down to aggregate user behavior (specifically, how often do users interact positively with your mailings) and your sending patterns (how much email do you send).

When adding new users to your mailing list, it's very important that users expect to begin receiving email from you. When users are surprised that they're receiving email from you, or don't remember signing up, they're more likely to mark your emails as spam -- and that tells mailbox providers that your email should be put in the spam folder rather than the inbox.

For the most part, mailbox providers aren't looking at the content of emails to determine what's spam and what's legitimate mail. Instead, each mailbox provider tracks how their users are interacting with your emails. Do they open? Do they spend time reading? Do they click? These are all positive signs of engagement that tell mailbox providers that your users like receiving your email, want to continue, and future emails from you should be put into the inbox and not the spam folders.

On the other hand, if users continually ignore your emails, mark them as spam, delete them without reading, or otherwise aren't interacting with your emails, these could all be signs that a user doesn't want to receive email from you anymore, and mailbox providers are more likely to mark future emails from you as spam.

Even users who signed up for your emails a long time ago might become disengaged and stop reading and interacting with your emails eventually. Some users might stop because their email address is no longer active; others might stop because they're no longer interested in emails from you.

It's important to unsubscribe users who haven't interacted with your emails in a long time. Mailbox providers notice this in multiple ways -- first, mailing users who don't interact with your emails makes your overall engagement metrics (open rate, click rate, action rate) much lower. This makes it look like your users aren't very interested in receiving your emails.

Mailbox providers also reclaim long-inactive email addresses and use them as "spam traps" in order to detect spammers who sign up users who definitely didn't opt-in, or senders who never remove inactive users from their mailing lists. Emailing a spam trap will seriously harm your sender reputation. By regularly unsubscribing users who haven't engaged with your emails recently, you can avoid falling into these spam traps.

What is a mailbox provider?

A mailbox provider is any email service that manages an inbox for its users, like Gmail, Yahoo, Microsoft, Comcast, among others.

Different mailbox providers have different means of identifying and classifying incoming email messages as spam.

How do I know how good my reputation is with a mailbox provider?

Some mailbox providers offer a dashboard that will allow you to see general stats, like Google Postmaster or Microsoft SNDS. Signing up for these services and monitoring them can help you identify potential problems and areas of improvement.

If you're seeing elevated delays or bounces at a mailbox provider, that's a sign that your reputation with that mailbox may be in trouble and further action may be needed.

How can I improve my sender reputation with a mailbox provider?

In addition to following the recommended advice given by Google Postmaster and Microsoft SNDS, we have some general advice.

Only send to users who actually opted in to receive your email

Ensure that you're sending email to users who actually opted in. Not only is this an ActionKit requirement, but it's good general practice. Users who did not opt in to receive your email are much more likely to mark your email as spam (which hurts you) and generally won't be positively engaged with your mailings.

Get users to engage with your email

Sending messages that engage your users will get them to open, read, and click your emails -- these are all positive signs to mailbox providers that users want to receive your emails and that your emails should be put in their inbox and not to spam.

Avoid clickbait, instead opting for compelling, timely content that's highly clickable and ideally has a clear theory of change.

A common mistake many campaigners make is over-relying on fundraisers, which generally have very low click rates. Make sure you're mixing up your content so your users don't disengage -- you can always redirect to a fundraising page from a survey.

Mail steadily and don't overwhelm users

Sending at a regular or semi-regular cadence is good because if it's been a long time since users have heard from you, they might forget having signed up. On the other hand, if you email too frequently, there's no way your users are going to be able to interact with them all, dragging down your engagement metrics.

Open and click rates aren't vanity metrics -- they're used by mailbox providers to determine what's spam and what's worthy of the inbox.

Aim for sending about once a week -- meaning your users should hear from you at least that often. We've often seen smaller organizations who go silent for very long periods of time (months or more) attempt to restart mailing, and mailbox providers identify this spike in sending volume as spam.

Drop inactive users

Inactive users by definition aren't engaging with your email. If someone hasn't opened, clicked, or otherwise interacted with any of your emails in a very long time, the chances of them doing so suddenly are vanishingly low.

By all means try to re-activate users with a re-engagement program, but keep expectations tempered -- very few users reactivate even after two tries.

If you don't unsubscribe inactive users, you run the risk of emailing "spam traps", which are email addresses that used to be active but the has been inactive for so long that the inbox provider has reclaimed them and now uses them to identify bulk senders who don't unsubscribe inactive users. When you email spam traps, you're telling mailbox providers that you don't unsubscribe inactives, and this seriously hurts your deliverability.

What tools does ActionKit offer me to help monitor, manage, and improve my deliverability and inbox placement?

Tools to help improve sender reputation through better list hygiene

  • Adopting a re-engagement process is one of the most impactful actions you can take to improve your sender reputation, because when you stop emailing inactive users, your overall open and click rates go up. These are the primary figures mailbox providers use to determine whether your email should land in the inbox or spam (or other folder like priority inbox, promotions, etc), and the better your metrics are, the better your placement.
  • Exclude typo domains: ActionKit can help you avoid sending to typo email addresses (which harms your sender reputation) by automatically excluding typo domains from your sends. This is enabled by default, and is configurable on the Configuration screen.
  • Soft Bounce Handling: unsubscribing users who repeatedly soft bounce because their inboxes are full (and never get cleared) can help your sender reputation, and we can help you manage this process automatically.

Tools to help improve deliverability through better targeting

  • When you target your mailings with ScorePool, you can send to your best, most active users -- and avoid sending to the least active users, improving your overall engagement rates.
  • Keeping an eye on the targeting summaries for your mailings lets you know what the baseline level of engagement is for your mailing list. Pay attention to this -- it can help you avoid targeting mistakes!

Miscellaneous tools to improve deliverability

  • Although it might sound a little silly, rate limiting your mailings is much better for your sender reputation than getting throttled by recipient mailbox providers.
  • Using the Order by Engagement Level feature to deliver to your most-engaged users first and your least-engaged users last can help improve inbox placement.
  • ActionKit excludes machine (fake) opens from our open rate calculations since Apple released Mail Privacy Protection in September 2021. Excluding these machine opens means that your reported open rates since then will be lower, but that's actually a good thing, because it means they aren't falsely inflated.

Tools to help monitor your deliverability and inbox placement

  • ActionKit's Daily Open Rates by Mailbox Provider report lets you keep an eye on the trends of how mailbox providers are placing your emails, or if you're possibly blocked.
  • ActionKit's Bounce/Delay Metrics screen pulls delivery data like delays and bounces from our upstream email provider SparkPost. This can tell you whether your mailing has an unusual number of bounces and/or delays.
  • Configuring a seed list to use with a deliverability monitoring service can help you gain greater insights into inbox placement.

What is a good open/click rate to aim for?

Every organization has a different list, whose composition will vary a great deal based on age, list size, level of activity, issue, and so forth. Because of this, it's extremely difficult to definitively say what's good and what's bad -- the important thing is looking at overall trends for your organization.

Generally, organizations with smaller list sizes have much higher levels of engagement.

Engagement tends to be a self-reinforcing feedback loop, though -- the higher your open and click rates, the more this is a signal for mailbox providers to place your emails in the inbox rather than the spam folder. Some mailbox providers and email clients also have higher-visibility/priority inboxes, or lower-visibility inboxes like "promotions", which can harm your open and click rates.

Generally, if you're sending compelling, relevant content on a regular basis to your supporters and regularly unsubscribing inactive users, you won't need to worry.

Organizations that regularly run tests to see what engages their list tend to have better open, click, and engagement rates. Test, re-test, and pay attention to what works!

What is a good spam rate to aim for?

While it's harder to give a specific answer for the right open/click rates, spam rates are easier. Your spam complaint rates should be as low as possible.

Google's recommendation in Postmaster Tools is to "Keep spam rates reported in Postmaster Tools below 0.10% and avoid ever reaching a spam rate of 0.30% or higher".

This is a good general rule, but keep in mind Gmail does not report spam complaints to us (or anyone), so it's important to check Postmaster Tools to confirm that your sends are below the 0.10% threshold. This also means that the spam complaint rates reported in ActionKit aren't counting complaints at Gmail.

What is an email warm-up?

An email warm-up is used to slowly increase your mailing volume over time. Warm-ups are used to avoid large spikes in sending volume that are typically associated with spammers. They're also about putting your best foot forward with mailbox providers, establishing yourself as a new sender.

SparkPost recommends warming up your list over the course of 30 days, beginning with your most active subscribers and gradually expanding your audience to less-active subscribers.

We strongly advise that you closely follow the recommended guidelines for warm-ups. We've seen many organizations ignore the guidelines in an attempt to "complete the warm-up more quickly" and pay the price in the result of worse deliverability and inbox placement.

By contrast, organizations that follow the warm-up guidelines (and our other advice on email deliverability) tend to have few problems with delivery or inbox placement.

What content should I send during an email warm-up?

During an email warm-up, your goal is to demonstrate to mailbox providers that your recipients love receiving your email, love opening it, love reading it, and love clicking it.

The best ways you can do that is by sending your best, most engaging content to your most engaged supporters.

A common mistake campaigners make is sending fundraisers during the warm-up process. We get it -- fundraising is a campaign necessity and you don't want to leave money on the table. However, the warm-up process is a risky time to fundraise and can seriously hurt your long-term deliverability and inbox placement.

This is because fundraisers are almost always less engaging than other forms of content -- they get fewer opens and fewer clicks. On the other hand, other types of content like surveys and videos tend to get much more engagement, showing mailbox providers that users want to see and interact with your emails.

Remember, you can always redirect to a fundraising page from your survey pages.

Why is it important to have a steady sending volume and cadence?

Having a steady sending volume and cadence is important for several reasons.

First, spammers are opportunistic, sending as many messages as possible before they get blocked by mailbox proiders. If you sent nothing for weeks, or you suddenly send to a lot more users than before, mailbox providers might think you are sending spam. Remember, they're not looking at the content of your messages -- they're looking at your sending patterns (how often you send, how many users you typically send to) and your users' engagement patterns (are users opening, reading, and clicking?)

While you might think "it's only been a few weeks since I last sent", or "don't the mailbox providers know I'm not a spammer? I've been sending for years", mailbox providers see it differently. A few weeks with no contact is a long time, and many mailbox providers don't tend to look very far back -- your recent sending history is much more important.

On the other end, how do your users view your email? If they haven't heard from you in a long time, they might have actually forgotten they signed up to receive email from you, or thought they unsubscribed.

It might sound funny that a user would forget they signed up for email from you, but it really does happen – and you risk them marking your emails as spam, which will hurt your future deliverability.

Finally, some number of email addresses will become invalid over time. This is normal, expected, and a regular part of a list's churn. Personal email addresses tend to last longer than student or work email addresses, but no email addresses last forever.

When an email address becomes invalid, it will become undeliverable, and the first time you send to that user, they will bounce and be removed from your list. That's good and fine on its own, but if you haven't sent in a long time, you'll have a lot more bounces all at once, which looks bad to mailbox providers who might think you've acquired a bunch of invalid addresses from some disreputable source instead of just not having mailed your supporters in a long time.

This is much worse than having the exact same number of bounces and unsubscribes spread out over several mailings.

What do I do if we haven't sent in a long time?

If you haven't sent in a long time, it's important to NOT pick back up at your previous sending volume immediately. If you do, you're likely to see higher than usual delays and bounces. See Why is it important to have a steady sending volume and cadence? and Bounce/Delay Metrics for more.

Instead, you'll want to slowly increase your sending volume over time. If it's been long enough, you're effectively starting at zero, so tread carefully, lead with your best, most engaged users, and go slowly. You'll have the most success if you treat this process like a warm-up, because it is.

You can expect higher numbers of bounces when you first begin sending, because if you haven't sent in a long time these users will not have bounced on their own. This can hurt your deliverability and sender reputation right as you're trying to get re-established. If you're not already enrolled in ScorePool, we highly recommend enrolling; when enrolled, undeliverable users will automatically be unsubscribed and you'll see a lot fewer bounces than you otherwise would.

What should I do if I see sudden list growth?

Most list growth isn't a cause for concern, it's a cause for celebration. In some cases though, you'll want to be careful.

Sometimes spambots will sign up to your list en masse for a variety of reasons. ActionKit has several anti-spam tools you can use to prevent comment spam and spambots from joining your list. But if they're already on your list, you'll want to identify these spambots and unsubscribe them via import.

What if you have a large influx of real users, though? If your number of subscribers has increased significantly, say 25% or more in a very short time, you'll want to be careful when emailing these users.

Remember, a significant increase in sending volume in a short period of time is how spammers operate -- and one way that mailbox providers detect spam. So even though you might be doing everything right, the mailbox providers don't know that these users signed up for your list legitimately; it may look like spam to them.

To prevent being mis-labeled as spam, slowly increase your sending volume over time. This should be balanced with the competing priority of activating your new users -- new users generally being the most active and engaged.

Some groups that have experienced rapid growth in a short period of time have segmented different parts of their list into "cohorts" that are treated separately because they tend to have different user behavior and patterns of engagement than the rest of the list.

Above all, experiment and find out what works for your list!

Targeting your mailings with ScorePool

ScorePool is a free, optional system that scores each of your subscribed users to predict how likely they are to engage via email in the next 30 days, based on their activity across all participating ActionKit clients. You can use ScorePool as a way to target users that are most likely to engage, or to exclude users that are unlikely to engage.

ScorePool works by aggregating data on opens, clicks, actions, and other similar data for every end user across all participating ActionKit clients into a shared data pool and collapsing all of this information into a single score used to predict how likely a user is to engage with a future mailing.

The data used by ScorePool does not include identifying personal information like plain-text emails, names, or addresses. The detailed activity data in the pool is kept private, anonymous, and internal; only the anonymized generated scores are shared.

You have to opt in to benefit from ScorePool. There is no cost to join ScorePool, but you do need to agree to permit the use of your anonymized activity data.

In addition to being able to target mailings based on ScorePool, organizations enrolled in ScorePool tend to see 20% fewer bounces than non-participating organizations. This is a huge improvement to list hygiene for participating users and can help improve inbox placement and deliverability.

Daily Open Rate by Mailbox Provider

ActionKit has a built-in report, Daily Open Rate by Mailbox Provider, which shows the open rates for your sent mailings, grouped by the largest mailbox providers.

This report can be a helpful guide to understanding the broader trends of your recent open rates. If you see any sudden dips, or notice that the open rates for one mailbox provider are very low compared to other providers, you might be blocked at that provider.

This report can be especially useful in spotting small deliverability problems before they become major problems; see also the Bounce/Delay Metrics screen for more.

Are some Mailbox Providers more difficult than others?

Every mailbox provider has their own set of rules that determine whether your email lands in the inbox or spam. Some providers are more strict than others in some areas.

Microsoft is generally considered the most strict, frequently blocking senders with generally good sending practices. They're especially punitive against senders who send more than once per day.

Comcast is also very strict, and is especially punitive against senders that they judge to be sending too quickly.

With these mailbox providers, be especially cautious with your sending volume and speed -- see using rate limiting and why it's important to have a steady sending volume and cadence.

In February 2024, Google and Yahoo introduced a new set of guidelines for bulk senders. ActionKit handles most of these requirements by default; the main takeaway here is to keep your spam rates below 0.10%. This is good advice anyway, and was the standard prior to February 2024; the new guidelines merely make that explicit.

How re-engagement can help

The main benefit of re-engagement is in eventually removing inactive users from your standard mailing stream. This helps because mailbox providers pay attention to how many of your users are inactive; if you're emailing people who never engage with your email and you're not eventually unsubscribing them, that hurts your reputation.

By setting up a re-engagement program, you'll increase your open rates and potentially improve your inbox placement.

Read more about how to set up re-engagement, which allows you to automate this process.

How re-engagement can hurt (if misconfigured)

A poorly-configured re-engagement program can harm your overall deliverability efforts. Here are our recommendations:

  • Don't mail inactive users more than twice. You won't reactivate very many users -- we've seen steeply diminishing returns after each additional contact.
  • Don't wait too long to mail your inactive users. If your re-engagement mailers are configured to send to users after even a few months of not having heard from you, it will appear to them as though you've just added them to your list, steeply increasing your chances of bouncing, mailing spam traps, and being marked as spam. We recommend sending a re-engagement email on the day they're added to the re-engagement list, and one week after.
  • If you have a lot of users on your re-engagement list, don't email them all at once. This is also true if you're setting up a new re-engagement program. You don't want to email all of your inactives at once, spread out your sends over time.
  • Don't give past donors a lifetime pass. Sometimes, we've seen overly-broad re-engagement queries for who's considered active define anyone who's donated within the last several years as active. That's too broad, because although donors are valuable to keep on your list, if they've stopped engaging, they're almost certainly no longer active.
  • Users who get added to the re-engagement list but aren't unsubscribed outright and eventually come back on their own might be inappropriately excluded from your normal email stream. (This is rare, but it happens)

If your re-engagement program is misconfigured, this can harm your overall deliverability and inbox placement because emailing your re-engagement list is emailing your least-engaged users. This can hurt because mailbox providers are looking at your overall engagement level -- and they don't know that you're trying to re-engage users. To mailbox providers, your re-engagement attempts might look like you're mailing users who haven't shown any signs of life in months.

How do I know if I'm blocked at a given mailbox provider?

ActionKit's built-in report Daily Open Rate by Mailbox Provider shows the open rates for your sent mailings, grouped by the largest mailbox providers.

Many things can impact open rates. Generally, your mailing's targeting has the greatest impact on its engagement; when you mail active users, you're going to see higher levels of engagement. When you mail inactive users, you'll see fewer signs of life.

Mailing relevant content to users is another critical piece -- users are much more likely to engage with content that is relevant to their interests. You can keep track of users' interests with tags or user groups to use later in reports and mailing targeting.

Looking at a mailing's targeting summary can tell you whether the overall composition of your audience is different as compared to your overall list or a specific mailing.

Assuming that your targeting hasn't changed, if your engagement is down, it's worth checking the Bounce/Delay Metrics screen screen, which can tell you whether you're seeing elevated delays or bounces at a specific mailbox provider.

When a mailbox provider is delaying delivery for your messages (usually because you're sending too much volume, or too quickly, or both), open rates will be down for that provider since some users may not have received your email yet, or if it has been delivered, it might have been delivered later than is convenient for your users, or it might have been delivered to a "promotions" (or similar) folder, or even to spam.

Delays themselves aren't signs that you've been blocked, but they can often be an early warning sign to change up your sending patterns before you get blocked.

If you're seeing elevated bounces and the Bounce/Delay metrics screen says your emails were rejected, or that you're on a block list, you're blocked and need to take action to get unblocked.

I'm blocked at a mailbox provider, help!

It's important to reach out early when you notice a deliverability problem like being blocked at a mailbox provider. If you don't take action, you risk making the problem worse in the interim.

Please reach out to support for help. We can investigate whether you're blocked, usually help get you unblocked, and make recommendations on how to keep you from getting blocked again in the future.

When you're blocked at a mailbox provider, we'll ask them to unblock you, commonly called "mitigation". In many cases, they will. But there's no guarantee they lift the block. Lifting the block can also take time -- sometimes several days -- and we don't always hear back on whether they've granted the mitigation request.

If you get blocked, this is a sign from that mailbox provider that they think you're spamming. So it's important that you change your sending patterns at this mailbox provider in order to put your best foot forward. That can include:

Finally, keep in mind that if you get blocked a second time, most mailbox providers won't give you a second chance. Be extremely careful and be sure to follow our recommendations.

When you're blocked, what we'll check first

If you think you've been blocked, let us know via support. We'll confirm by checking your open rates by mailbox provider over time, your recent mailings' delays and bounces, and your recent sending history.

If you've recently made any large imports or targeting changes to your emails, please let us know -- this can help us more quickly identify whether these have contributed to your current deliverability problems.

Sending us the recent mailings you're concerned about and what you've already looked at can help us more quickly diagnose what might have been the root cause.

We'll look at that, plus aggregate analytics of your broader sending patterns in order to determine when the blocks started and what might have been the original causes.

If appropriate, we can request mitigation on your behalf to get you unblocked at a mailbox provider.

What we can't help with

Not every deliverability situation can be resolved through mitigation. In some cases, it's not appropriate, and it's not to be used as a "get out of jail free" card.

If we think there are approaches other than mitigation that would be more appropriate, we'll suggest those, especially because mitigation is something that not all mailbox providers offer, and even those that do are unlikely to grant it more than once.

Mailing Targeting Summaries

Once you've built your mailing set, we'll calculate statistics about those recipients so you can see how they compare to subscribed users on your list who have received mailings recently.

We'll show a brief summary of this information on the Proof and Send screen, and more detailed information on the Compare Mailing Set screen.

This information can be especially useful when comparing the relative performance of your mailings. Sometimes, a campaigner might make a targeting mistake and accidentally exclude all actives instead of including them -- and spotting a mistake like this is much easier when you pay close attention to the "Average Activity" stat for your mailings.

More generally, these stats are useful as a sense of the overall activity levels of your mailing list.

Order by Engagement Level (Scorepool)

The Order by Engagement Level option on the Mailing targeting screen (under Limits and Ordering) can potentially help improve deliverability and inbox placement by delivering to your most-engaged users first and your least-engaged users last.

This option is only available for clients enrolled in ScorePool, a free, opt-in system that assesses how likely your users are to engage with your mailings in the next 30 days.

Order by Engagement Level may be especially effective when paired with rate limiting. By delivering mail to less-engaged users after your most engaged users have already opened and clicked, you might get better inbox placement than if you had sent without it.

Warning

Use caution if you use Order by Engagement Level when conducting an A/B test with a limit -- your test results will show high levels of engagement because you're sending to your most engaged users, but when you send the winning subject line to your full list, it almost certainly won't perform to expectations, because all of your most engaged users have already been emailed.

In other words: when conducting an A/B test, try to keep your audience random.

Checking the Bounce/Delay Metrics screen

The Bounce/Delay Metrics screen pulls data from SparkPost to determine whether your mailing has an unusually high number of bounces and/or delays.

Pay attention to this and our warnings here -- you should treat this as an early warning system for broader deliverability issues, and take action before delays turn into blocks.

Why are my mailings being delayed?

Delays can happen if a mailbox provider decides that you're sending too much email too fast relative to your sender reputation with them. This is especially common and expected during the warm-up period when you are first establishing your sender reputation. Remember that your sender reputation is determined by each mailbox provider separately. For example, you may have a good reputation with Gmail but a poor reputation with Comcast.

Note

If you're not seeing delays, then you're sending at a speed and volume that's consistent with your sender reputation for that mailbox provider. If you are seeing delays at a specific mailbox provider, this is their way of telling you that you are sending either too much email for your sender reputation, or too quickly for your sender reputation, and that you should slow down, send less volume, or both.

Some mailbox providers are also more sensitive to sending speed or volume. Typically, smaller mailbox providers tend to prefer that you send more slowly, but Comcast and Microsoft are more sensitive and will often penalize fast senders. We've also seen Microsoft penalize senders who send multiple emails per day.

If your mailing has very few recipients, a fast sending speed may not be a problem, but could arise with more recipients. For example, you may not see delays if your mailing has only 1,000 recipients but might see delays with 10,000 or 100,000 recipients, all other factors being equal.

How fast are my mailings sending?

When viewing a sent mailing's Mailing Report screen, look for the "Sending Speed" metric just above Content and Targeting. This will tell you roughly how quickly your mailings are sending. This is an overall stat, and isn't broken down by mailbox provider.

Keep in mind that the reported sending speed includes a few minutes of server start-up time, so the metric shown is a slight undercount of the actual speed. This is more noticeable when your mailing has fewer recipients.

You should take care to look at each individual mailing, because unless you're setting a rate limit (Max emails/sec on the Targeting screen), each mailing will be sent at as quickly as possible. (The actual speed could vary widely based on a variety of factors beyond the scope of this document.) In other words, looking at a single mailing isn't enough to definitively judge the baseline sending speed of the rest of your mailings.

What can I do about mailing delays?

Before taking action, it's important to identify whether your mailing delays are the result of sending too much mailing volume, sending too quickly, or both.

If your mailing delays are the result of sending too much mailing volume, the remedy is to back off and send less volume.

If your mailing delays are the result of sending mail too quickly, the remedy is to use rate limiting to slow your sending speed down instead of having the mailbox provider slow your sending speed down for you.

If your mailing delays are the result of sending too much mailing volume, using rate limiting may be somewhat effective, but not as effective as backing off and sending less volume.

If your mailing delays are the result of sending mail too quickly, sending to a smaller audience may be somewhat effective, but you're likely to contiue to see delays.

If your sending volume has changed recently, or you've had significant list growth, it's important to slowly onramp these new members onto your list over the course of several days or weeks. Mailbox providers look skeptically on senders who have dramatic spikes in sending volume, because that's typically how spammers behave.

If your sending volume isn't the root cause, rate limiting your mailings may help. See Using rate limiting.

How to use Rate Limiting

If you're seeing mailing delays as the result of sending mail too quickly, slowing your sending speed can be an effective way to avoid having your delivery delayed by a mailbox provider.

In these cases we generally recommend sending as slowly as you can. We often see clients using rate limiting with such high limits that it's ineffective.

Of course, there are legitimate campaign reasons to send an email quickly -- we understand there's a need to get your emails to supporters in a timely fashion. There's no one-size-fits-all answer to "how fast should I send", in part because it's a balance between how many recipients you have, your campaign needs, your sender reputation, and how timely/evergreen your mailing is.

And if you're not seeing delays, you can ease up on rate limiting with the goal of eventually removing it entirely.

With all that in mind, a good way to determine how slow to actually go is to look at the recorded sending speed of mailings where you saw delays and aim a good deal slower than that.

Mailings with more recipients will take longer to send. For example, if your mailing has 10,000 users in its targeting, this will complete faster than a mailing with 100,000 users.

How long will my mailing take to finish sending?

This mailing would take about 33 minutes to complete.

To reiterate: when experiencing delays due to sending too quickly, we recommend sending as slowly as you can. ActionKit supports sending as slowly as 2 emails per second, which sounds very slow but if your sending size isn't that large, is probably fine:

At 2 emails/s, a mailing to 10,000 recipients finishes in 1 hour 23 minutes.

Depending on the number of recipients you have, you could likely get away with sending a good deal faster than that -- perhaps 20 emails/s, which for a mailing to 100,000 recipients also completes in 1 hour 23 minutes.

Generally, if you're seeing delays on the Bounce/Delay Metrics screen and the reason says to slow down, then slowing down can help.

Don't Segment out Problem Domains

Finally, we generally do NOT recommend segmenting out your list to a single problem domain/mailbox provider when rate limiting. This is a common mistake we see clients make -- after being delayed at Comcast for example, you decide to send a rate limited email to your Comcast subscribers.

While tempting because it ensures your mail to recipients at other mailbox providers doesn't get slowed down by rate limiting, unfortunately this won't be as effective. Consider an example mailing of 10,000 recipients, of which 10% are at Comcast. Assuming these subscribers are randomly distributed and the mailing is rate limited to 10/s, this has an effective rate of 1/s at Comcast. Contrast this with a separate mailing of 1,000 Comcast recipients rate limited at 2/s -- twice as fast.

What are mailing bounces?

When you send email to an address that is permanently undeliverable, the recipient mailbox provider will reject your email since it can't be delivered. This is called a hard bounce.

There are many ways an email address might be permanently undeliverable, the most common being that an email address is invalid, is a typo, or was once valid but is now disabled.

ActionKit automatically unsubscribes users that hard bounce when you email them.

If you're seeing elevated bounces on the Bounce/Delay metrics screen, this may be a sign that your emails are getting rejected by the recipient mailbox provider, usually because they think you are sending unsolicited bulk email (spam).

Most commonly, a recipient mailbox provider will block delivery as a last resort. In most cases, mailbox providers will delay messages many times first as a way of saying "slow down". If this warning goes unheeded, those delays may turn into bounces. Read more about our Bounce/Delay metrics screen.

Generaaly, ActionKit does not record soft bounces in the database because most soft bounces are transient errors that may be recoverable later. Contrast this with hard bounces, which are permanent. In one specific case however, ActionKit does record soft bounces -- those associated with a user's mailbox being full.

A user's inbox being full isn't a problem generally, and isn't always a permanent problem. Users who are still active will clear their inboxes eventually, allowing them to receive email again.

But some portion of users with "mailbox full" soft bounces are inactive and won't ever clear their inboxes. Unsubscribing these users who repeatedly bounce can help improve your deliverability because it shows mailbox providers that you don't repeatedly send to users who can't receive mail and are practicing good list hygiene. Read more about soft bounce handling.

How does Apple's Mail Privacy Protection (MPP) program impact ActionKit?

In September 2021, Apple released Mail Privacy Protection (MPP), billed as a privacy feature for their users. This feature makes it so ActionKit (and every other sender) is unable to confirm whether an MPP-enrolled user has actually opened your email or not.

Email opens are tracked by whether an image (often called a tracking pixel) is loaded by the recipient. MPP confuses this by loading all images for all emails, whether the email was opened or not.

Because we can't confirm that users enrolled in MPP have actually opened your email, we don't count these opens in your Open Rate reporting statistics. As a result, open rates have decreased significantly since MPP's release -- but the key takeaway here is that these rates are not inflated.

Open rates before and after the introduction of MPP

On the left are open rates before the introduction of MPP. On the right side of the red line are what open rates would be after the introduction of MPP if we didn't filter out the fake opens. In other words: ActionKit's underreporting of open rates here is a good thing.

Sidebar: Some email clients don't load images by default, and for users on these email clients, they may not record an open. This has always been the case, and open rates have always been an undercount as a result.

Opens being an undercount means it's especially important to have other metrics (like clicks) factored into your determination of what's considered an "active" user for purposes of mailing targeting and re-engagement.

ActionKit does not "infer" opens from clicks or actions, so it's not unexpected that a user might have a click on a mailing but no open, especially if this user is enrolled in MPP.

Can I outsmart MPP?

Although it might be tempting to look for patterns in machine opens, like ignoring machine opens within seconds of receipt, counting multiple machine opens as signs of a real open, or scrutinizing the timing of machine opens, these aren't reliable signals.

A machine open will get recorded by an iOS device connected to power and/or WiFi for an MPP-enrolled user. A user might have more than one such device, and may be connected to power and/or WiFi whether or not the device is actually in use.

In other words: a machine open may be recorded when the device is not actually in use by the user.

The only reliable signal from a user with a recent machine open is that they are not a spam trap.

MPP Technical Details

A machine open for these users will be recorded in the core_open_raw table, but it's important to note that this shouldn't be considered a sign of activity.

core_open_raw contains machine opens AND real opens alike.

Anyone with an iOS device can be enrolled in MPP regardless of what email address they have, though users with Apple domains (including icloud.com, mac.com, me.com) may be over-represented among MPP enrollees.