While many marketers might bemoan the effort it takes to reach the inbox, the hurdles put in place by mailbox providers are designed not to block legitimate senders but to keep spammers and harmful content out of their users’ inboxes.
To identify the “good” from the “bad,” mailbox providers judge an incoming sender’s reputation when determining what email to allow or block from their inboxes. A sender’s reputation is calculated using a variety of metrics, and each mailbox provider has their own formula to judge whether or not to accept incoming mail and where to filter it.
Marketers who are aiming for the inbox use Return Path’s Sender Score to discover and track their sender reputation and learn how mailbox providers perceive their messages. Similar to a credit score, Sender Score is an evaluation of your sending practices compared to other senders to determine the quality of your messages and your email program. A marketer’s Sender Score is expressed as a number between 0 and 100. Sender Score uses data points and reputation formulas similar to what mailbox providers use to give a relatively accurate representation of how mailbox providers view your email.
In the report below, we will discuss some of the factors that go into both Sender Score and mailbox providers’ reputation calculations. In addition, we’ll take a look at the impact those elements have on how you are evaluated and how that judgment impacts your ability to reach your subscribers.
When we first conducted this benchmark back in 2012, we found that 60 percent of messages sent came from the least reputable senders—those with a Sender Score of 10 or below. By contrast, the best senders—scoring above 90—accounted for less than six percent of total volume.
Today, 36 percent of sent mail comes from senders scoring above 90, while the mail from the least reputable senders has dropped to 25 percent.
This shift in volume distribution by sender reputation is largely due to mailbox providers increasingly blocking email from senders with spammy behavior. These stricter protocols for reaching the inbox has made it increasingly difficult—and expensive— to successfully send spam.
While each mailbox provider has their own formula which assigns different weights and values to various components of sender reputation, the factors they include are generally the same. Three key metrics that factor into sender reputation are complaint rate, unknown user rate, and spam trap count.
These metrics are included in reputation calculations because they provide insight into a sender’s mailing practices. Those with low complaint and unknown rates and few spam trap hits are seen as reputable senders and receive a higher reputation score. As a result, inbox placement tends to be higher for these senders.
A complaint is generated when a mailbox user marks a message as junk or spam. A high complaint rate signals to mailbox providers that the sender’s content is unwelcome and should be sent to spam.
Senders scoring above 90 were the only ones who managed to maintain a complaint rate below one percent. Those with a Sender Score between 11 and 90 had an average complaint rate ranging from between 2.8 percent and 3.6 percent, while senders scoring below 10 had an alarmingly high average complaint rate of 7.4 percent.
Senders scoring between 71 and 90 produced a lower average complaint rate compared to last year, while senders between zero and 60 saw much higher complaints.
An unknown user is an email address that never existed, has been terminated by the mailbox provider, or was abandoned by the mailbox user. Mailbox providers will return a hard bounce code (5xx) indicating when email is sent to an unknown user. Senders who not only retain unknown users but send to a high percentage of them are perceived by mailbox providers as having poor list hygiene practices, which will impact their sender reputation.
Senders scoring above 90 tend to keep relatively clean lists, with average unknown rates of one percent. The second highest band (81-90) saw an average of 3 percent unknown users.
Senders in 2017 maintained a higher list quality than the previous year, with each band seeing a decrease in average unknown rate.
In most cases, marketers should remove addresses as soon as they receive a 5xx hard bounce code. To learn about the different types of bounces and how to process them, read the Email Marketer’s Guide to Bounce Processing.
Spam traps are email addresses that don’t belong to active users and are used to identify both spammers and senders with poor data quality practices. There are two types of spam traps: pristine and recycled. Pristine spam traps are created solely to capture bad mailers. These email addresses were never owned by a real person and therefore shouldn’t be on any marketer’s list. Recycled traps are addresses that were once held by a user, but have been abandoned and converted into spam traps.
Senders scoring above 90 were the most successful at keeping their lists clear of spam traps with an average of 0.36 spam traps hit. Senders scoring 50 and below saw an average of one or more spam trap hits, with those scoring between zero and 10 hitting an average of 7.5 spam traps.
Spam traps were more frequent in 2017 than in 2016, with each Sender Score band seeing an increase. Those scoring 60 and below had significant increases over their 2016 average spam trap count—often double or more.
To keep your program safe, make sure you understand the different types of spam traps and how to identify and remove them. Also, consider investing in a list validation service to help you identify and remove traps on your list and avoid acquiring new ones.
A sender’s delivered rate measures the percent of messages that aren’t bounced or rejected by mailbox providers’ gateway filters. The delivered rate does not show which folder the messages were then delivered to—primarily the inbox or the spam folder—but it does factor out any messages that have no hope of being seen by subscribers.
In 2017, senders scoring above 90 had an average delivered rate of 91 percent. Those who fell one band below (81-90) saw far less mail delivered, with an average of 68 percent. The remanding senders—those scoring between zero and 80—saw more than half of their messages rejected at the gateway.
2017 saw slightly lower average delivered rates by Sender Score band compared to 2016.
After passing the gateway, emails are subjected to additional filters, including both reputation-based filtering and engagement-based filtering. As discussed previously, each mailbox provider has their own unique formula for evaluating reputation, although the factors that weigh into this analysis are generally the same. As a result, senders may experience different inbox placement rates at different mailbox providers.
In the graph below, you can see the impact of these various filtering formulas decisions at Gmail, Microsoft, Yahoo, and AOL. While different, it’s worth noting that each mailbox provider’s inbox acceptance generally increases with a higher Sender Score and better sending practices.
Of the four mailbox major providers, inbox placement decisions at Yahoo, Microsoft, and Gmail show a strong correlation with Sender Score. Senders scoring below 60 see less than half their email placed in the inbox at these three providers. Yahoo is a little more lenient, allowing senders scoring as low as 81 to exceed the global inbox placement rate (80 percent). Meanwhile at Gmail, only senders with reputation scores above 90 see an average inbox placement rate above 80. Microsoft is the most difficult mailbox provider to reach, with a maximum average inbox placement rate of 72—achieved only by those with a Sender Score above 90.
The chart above shows that sender reputation filtering is still important in spam filtering algorithms at mailbox providers, as senders with a higher Sender Score generally receive better inbox placement, particularly at Yahoo, Microsoft, and Gmail. Marketers need to pay attention to the impact ignoring complaints and retaining spam traps and unknown users on their list can have on their reputation, inbox placement, and their ability to reach their subscribers.
However, a high Sender Score on its own doesn’t translate to higher inbox placement rates. Subscriber engagement, a mailbox provider’s own reputation calculations, and the content in the incoming message—none of which are included in Sender Score calculations—all factor into each mailbox provider’s final filtering determinations.
To help marketers evaluate their own email program against their peers, we also dug into Sender Score performance by industry. The chart below shows the average Sender Score and inbox placement rate for more than 17,000 commercial senders, broken out by industry. As only legitimate commercial senders are included, these numbers are a better reflection of how each industry is performing.
By industry, Sender Scores ranged from 81 (telecommunications) to 95 (apparel & accessories, flowers & gifts, kids & babies, and sporting goods), with the majority of industries scoring above 90. Interestingly, all industries either maintained or increased their average Sender Score in 2017, with the technology/software/internet industry seeing the greatest increase (11 points).
To conduct this study, Return Path analyzed over 6 trillion messages sent during 2017 from IP addresses whose Sender Score was calculated, and whose subscriber engagement and inbox placement data were available for analysis. In addition to Sender Score data, this report used data from the Return Path Reputation Network to track inbox placement rates across mailbox providers, and the Return Path Consumer Network to identify the Sender Score and inbox placement rates of more than 17,000 commercial senders. Data used for this report is aggregated and anonymized, and is not limited to Return Path clients. Sender Score is a free reputation calculation service and is available at senderscore.org.