Sales professionals, marketers, and recruiters live and die by two vital metrics when it comes to email: bounce rates and open rates.
Ultimately, it doesn’t matter how compelling an offer may be if an email fails to reach its target or they choose not to open it. Inaccurate email data can jeopardize sales, limit the reach of marketing campaigns, and let quality candidates slip through the cracks.
Email data accuracy is a major priority for ZoomInfo’s data engineering team and a core part of the ZoomInfo platform. No other B2B business intelligence product offers the same level of email data fidelity, thanks to our unique combination of sophisticated automation technologies and in-house human expertise.
In this post, we’ll explore how ZoomInfo treats email data to ensure that our customers have the most accurate, up-to-date information for their go-to-market strategies and campaigns.
Going beyond SMTP checks
Given the importance of accurate email data, it’s hardly surprising that there are dozens of products, tools, and services available to check the validity of email contacts. Services such as Bouncer, BriteVerify, DeBounce, and many others all promise quick and easy email verification.
Most email verification tools rely primarily on Simple Mail Transfer Protocol (SMTP) checks. SMTP is an email communication protocol that is used to send and receive email.
In a standard SMTP exchange, a mail client sends an email to an email server in a process known as an SMTP handshake. Next, that server sends the email to a mail transfer agent (MTA), which then identifies the recipient domain — the string that follows the “@” symbol in an email address. Finally, the MTA chooses a server on the recipient’s side and completes the exchange.
SMTP checks use variations of this process to verify email addresses. However, while simple SMTP checks can verify that an email address exists and receive incoming messages, there are many potential pitfalls when relying solely on SMTP checks.
For example, poorly or incorrectly configured mail servers can return false information about a specific contact, such as confirming an email address exists even if it doesn’t, which would result in a bounce. At scale, such inaccuracies can render entire email lists practically worthless; the average email list decays by approximately 3 percent every month, meaning almost one-third of an email list can stagnate in just one year.
At ZoomInfo, we’ve been continually refining our email verification process to provide a level of accuracy that no other platform or product can match. A huge part of those efforts was the acquisition of NeverBounce technology, which we began integrating into the ZoomInfo platform in March 2019.
NeverBounce uses a proprietary cleaning process with more than 20 steps — checking each email contact up to 75 times — to provide our customers with unmatched accuracy. Some of these processes include:
- Syntax checks to ensure a valid email address
- Distinguishing between role-specific email addresses, free email addresses, and email accounts associated with educational and/or governmental institutions
- Comparisons between known datasets to identify bad, defunct, and disposable email domains
- Verification that email domains exist, are valid, and refer to a specific email server
- Real-time connection queries to specified mail servers
- Machine-learning models that examine historical data for specific email contacts
NeverBounce doesn’t just verify individual email addresses. It can verify entire mail servers using a combination of advanced SMTP checks and processes to ensure user identification data is accurate, all in real time. This translates into significantly lower bounce rates, higher revenue, and more effective campaigns.
However, even advanced SMTP checks are still far from perfect. Some email domains appear to accept all incoming mail regardless of whether contacts are valid or not. This is largely due to security services, such as Mimecast, that help protect against threats like phishing attacks by obscuring whether the intended addresses were valid.
Because of such services, SMTP checks can only ever accurately validate 50 percent of business domains, regardless of provider. To overcome this limitation, ZoomInfo leverages unique, proprietary send-and-bounce data to check domains that cannot be verified by SMTP checks alone, which offers our customers significantly greater accuracy.
We also see the email verification process as a unique opportunity to provide additional value to our users. Every day, we send research surveys to more than 90,000 contacts asking for their feedback. This provides us with invaluable insight into potential features ZoomInfo customers may find useful and helps us validate email contacts in our database.
Accurate email data is critical to the value of the ZoomInfo platform, but it’s also a delicate balancing act in terms of privacy. Similar to our research surveys, we regularly send privacy notices to contacts to ensure that we’re only reaching out to individuals who have given us their express permission to be contacted. This helps us respect recipients’ privacy, ensure compliance with regional data usage directives, and further improve the accuracy of our email database.
Turning data into insights with machine learning
The steps above are an excellent starting point for our data engineering team, but it’s only the beginning. ZoomInfo’s sophisticated machine-learning technologies allow us to offer a level of precision that other products simply cannot match.
We analyze our vast repository of user data with thousands of data points to effectively train cutting-edge learning models to predict the likelihood of an accurate email. These models evaluate hundreds of data signals, such as company size and patterns of email activity. As the models evaluate more data over time using industry-leading, continual-learning technology, the more accurate the predictions become, creating a virtuous cycle of continual improvement and superior data fidelity.
ZoomInfo’s Contributory Network is another invaluable source of data for our machine-learning models. In exchange for access to ZoomInfo’s Community Edition, the 200,000+ members of our Contributory Network agree to allow our machine-learning technologies to analyze the email signatures of messages sent and received by their email accounts. Thanks to the data provided by this ever-growing network, we are able to verify more than 20 million email signatures every month, and update records for over 4 million individuals and 1 million companies in our database every single day.
Adding the human touch: ZoomInfo’s research team
Even the most sophisticated automation technologies can benefit from human expertise. ZoomInfo’s in-house team of more than 300 experienced researchers and data analysts spend almost half their time on data quality initiatives, such as verifying the information gathered and processed by our machine-learning models. They provide an additional layer of scrutiny and data validation to ensure our customers have the most accurate data possible.
The judgment and expertise of ZoomInfo’s researchers provides breadth and depth to our data that far exceeds that of other companies and platforms. It’s a critical advantage that strengthens our position as the industry leader in data accuracy. Our researchers are available to provide crucial customer and sales support, and frequently conduct custom research to cater to specific client requests.
Bringing technology, expertise, and vision together
ZoomInfo’s holistic approach to email accuracy and verification enables us to go far beyond rapid automated solutions. Our combination of sophisticated automation technologies and human expertise allows ZoomInfo to offer both scale and accuracy, rather than forcing our customers to choose between one or the other.
We’re constantly investing in our technologies and our team to provide even greater value to our customers. To learn more about ZoomInfo’s data quality and how we can help your team hit their number, check out the following case studies: