There is considerable market demand for Email Address Verification (EAV). This is reflected by the fact that numerous companies have specialized in this field. For example, Return Path, Inc., the largest email intelligence and deliverability service provider, which works closely with numerous ESPs and ISPs, has recognized this fact and has started to collaborate with a major provider of EAV in order to offer its customers a benefit in email marketing.
Wherever email addresses are processed, at least one syntax check takes place. EAV starts here, and extends it by domain-specific syntax rules, recognition of typos, DNS/WHOIS/HTTP-based checks to determine the operators of domains and the mail servers behind them. This is helpful to evaluate how useful or harmful an address is in context of email marketing (all types of spam traps, disposable/temporary, role and bot addresses would be harmful) and can also be used for various other kinds of segmentation, for example in B2B and B2C addresses. Finally the standard mail transfer protocol SMTP is used to check the status of the mailbox, whether an attempted mail delivery to that destination would succeed.
Real Life Examples of Bad Addresses
|firstname.lastname@example.org||Invalid top level domain|
|xyz.@gmx.de||Invalid mailbox name|
|email@example.com||Invalid Syntax (domain specific). The mailbox length for Gmail must be between 6-30 characters|
|firstname.lastname@example.org||Non-Existent Domain (DNS)|
|email@example.com||Domain exists, but can’t receive mails|
|firstname.lastname@example.org||Dead Domain Spamtrap|
|email@example.com||Dead Domain Spamtrap|
|firstname.lastname@example.org||Classic/pure Spamtrap (never assigned to a real user but receiving email)|
There are different views, standpoints and use cases on this subject. In general, one can distinguish between the view of the Email Marketer, the Email Service Provider (sender) and how it is seen by the receiving side (the ISPs and antispam organizations).
Email Marketers and Email Service Providers Perspective
From the standpoint of email marketing, the most important use case is the integration into online forms. Many addresses obtained through these forms are invalid because of typos made while putting in the email addresses. This is, for example, one of the main reasons why double opt in mails bounce. We see 2% to 5% DOI bounce rates if no EAV service is used on the online form. Handwritten email addresses generated on paper form, for instance sweepstake forms, generate even more typos and thus higher bounce rates. Many of the bounces that originate from typos can be avoided by EAV, because the end user gets real-time feedback and can correct the typos while one on the form. In addition, good EAV services can protect email marketers from harmful email addresses entered on forms. The promise is that at the end of the day you will have more subscribers and also a higher conversion rate, because problems around of fake signups, disposable addresses, typo domain spamtraps etc. can be prevented.
An Email Service Provider can offer EAV as a service to its customers in order to improve their email collection methods and the quality of data in general. It can also be used internally as a monitoring or deliverablity tool. With EAV you can detect spamtrap hits without having to run your own spamtrap network. The information gained by this technique is very helpful in identifying problematic customers and helping them to fix their problems related to unclear/poor permissions, enforce stronger list hygiene measures, remove bounces and bad list segments.
Whenever new addresses are uploaded into the system, a random sample of them can be automatically checked using EAV, and if certain thresholds are exceeded, further actions from simple alerts to deactivating the whole customers account can be taken. This prophylactic measure helps avoiding all kind of negative effects resulting in sending to bad addresses. A central deliverablity task is to decide when an address should be deactivated, after it has bounced. This is normally done by looking closely at the SMPT error messages (DSNs) of the particular sendout. With EAV one would have an alternative way to determine if the bounced address can or can not be delivered in principle. For example, bad newsletter content can easily result in a 5.7.1 hard bounces (security/policy violation) and lead to the unwanted deactivation of email addresses that are otherwise perfectly safe to be delivered to.
We as an ESP have developed our own EAV service (www.addresscheck.eu) and have been using it extensively to help our clients generate more and better email addresses on their online forms, and to protect ourselves from dangerous lists.
The Recipients (ISPs) and Anti-Spam Organizations Point of View
In fact, there is also a use case for the recipient side, namely as an anti-spam measure. Since a percentage of spam is generated from forged Envelope Sender addresses, some spam can be detected by checking whether forging resulted in an invalid address by using this technique. This is known as “Sender Address Verification” (or “Callback Verification“) and is supported by Postfix and Exim, among other mail servers. However, because of the expense, it is only used by small mail systems. The major German ISP 1&1 Internet SE (GMX.de/Web.de) applies only a “light” version of this by checking only the domain part of the Envelope Sender and blocking it if it cannot receive mail.
Besides this, however, in general EAV is viewed somehow dubious by the recipient side and the anti-spam organisations. One of the reasons for this is that it can be misused by spammers for listwashing and spamtrap washing. Chris Thompsen from Spamhaus, has written a comprehensive article here, where he weighs the pros and cons against each other and gives some recommendations if you want to run such a service. Mainly such a service should only be offered in the context of Permission Based Marketing. With a strong emphasis on permission. This means that addresses whose origin is unclear or are without permission for bulk advertising should not be verified.
*Please note this article relates to the following video: