CarePrecise works diligently to keep healthcare provider email inventory up-to-date and deliverable. As hard as we work, there are innumerable roadblocks to deliverability that rely on savvy email marketing that are, frankly, completely out of the control of the email supplier.
Especially since the COVID-19 pandemic, reaching healthcare providers by email has become increasingly difficult. Unscrupulous vendors of personal protective equipment (PPE), often overseas vendors with little understanding of proper email marketing techniques, launched a constant spew of spam email to physicians, PAs, NPs, and others desperate for PPE. According to reports from CarePrecise partners and customers, as many as thirty spam messages per physician were filtered every day for months, protecting clinicians from stress and wasted time. Still, many got through, angering practitioners and causing further tightening of spam filters. (Note that CarePrecise is careful to vet email buyers by checking the IP address of their inquiries, and running WHOIS and reputation checks against the IP and message domain. We do not want to participate in spamming.)
Even before the pandemic-related tightening – much of it automated, and not under human control – email servers hosting medical email systems were working to reduce vendor email to clinicians. The filters "learn," programmed to adapt as email traffic changes. With fewer spammy emails in the flow, the healthcare community can expect deliverability of legitimate marketing email to increase, perhaps back to previous levels. For now, email marketing to healthcare providers is a unique challenge. (Apple's new Mail Privacy Protection and Hide My Email schemes are also going to gum up the works. Read here about this new issue in email marketing.)
What can you do to convince the filters that your email message is worthy of delivery? Here we present some of the best practices in post-pandemic email marketing.
Getting Email Into the Mailbox
Email sending vendors typically stand by their determinations, but when compared side-by-side with another vendor, the results can vary quite a bit. Pre-flight verification is always a good idea, but the best use of this is to separate email addresses into different classes, or grades. It is wise to send only small batches every time, but especially those that result in "Valid Catch-all" and "Greylisted," as these may be more likely to trigger anti-spam filters, server rejection, and other server behaviors, resulting in undelivered email. Also, if your domain sends too many of these emails at one time your domain reputation may suffer. Even so, these steps cannot guarantee that a given email message from a given sender will "get through."
The art of sending email has much to do with crafting a message that doesn't look to a Bayesian filter like earlier email that may have turned out to be actual spam. If the message seems "salesy," if things like exclamation marks or emojis are over-used (especially when they appear in the subject line), if text runs too long before a full paragraph line break, if poor spelling, punctuation, grammar, or other linguistic characteristics suggest a poor grasp of business tone, or a mismatch with the interests of the ISP's general population... and we could go on. This is the reason for testing small batches.
After running a pre-flight verification (with an email verification service that returns useful reasons for undelivered email), a small number of email addresses from each grade, or class, of email is sent and the bounce report carefully examined for the reasons. In each class, if a high number of "Greylisted" results are obtained, the next group of emails in the class should tighten up on the suggestions for getting through spam filters. If a lower class of email continues to test poorly, accept the fact that aspects of your - necessary as they are to your message - will not penetrate. Send no large flights of this class, or you risk the reputation of your domain, which may affect even your ability to send normal business emails.
What follows are good practices that will result in more of your emails reaching mailboxes, and get results for your campaigns.
- Make sure to comply with the U.S. CAN SPAM Act. Be aware of, and conform to, the laws of your governments and those of your addressees.
- You are required by law to include an "Unsubscribe" capability, and you must follow through and remove unsubscribers expeditiously.
- Limit the number of emails that are sent per hour and per day.
- Ensure that the contents of the message does not read like spam (there are services for that).
- Use full paragraph line breaks (character return+linefeed) instead of <br> or <br /> line breaks. The single break may be invisible to some filters, and extremely long lines of text are indicators of computer-generated messages created by sloppy spammers.
- Verify that your DNS security settings are properly configured.
- Include the current date in the subject line or body of the email.
- Include the recipient's name in the body like "Dear Dr. Pendragon."
- Include other personalizations that make email message appear different - and more like personal communications - to machine readers.
- If you have repeated something, put one version of it into one email flight, and the other different version into another email flight, thus creating differentiation that can penetrate filters. Note that if you have URL links in your message to a page that states exactly the same thing as your email message, some filters are smart enough to know that this is a sales message, and a subset of those filters are set to reject every sales message except for those with certain inscrutable characteristics. In each of the smaller flights, mix up the recipient domains and differentiations so that each batch has a mix of looks.
- Use your natural voice and personality; humans get past filters better than robots.
- Check for proper spelling and grammar. Email messages that look like they might have come from a sloppy foreign sender are likely to get labeled as spam.
- Keep it short; if you find that you've repeated something, just don't. Forget the old school "Tell em, then tell em what you told em, them that you told em twice." That was then, this is email.
- Be wary of offering "Free" give-aways; this is a widely known filter trigger (and never put the word "free" in the subject or description line. The terms "no charge," "no cost," and even "affordable" are almost equally bad. Instead, offer something that is "understood" to be free without actually saying it, like a whitepaper, or a webinar. As of this time, you can use the terminology of action relatively safely, like "we can do it now" or "let us try it," but resist the temptation to include "no obligation." The wrong language is responsible for many blocked sends.
- If you include graphics in your emails, such as a logo, use different sizes and aspect ratios, as well as different image file names, to different blocks of email addresses. Differentiation here means email messages that do not look nearly identical to a Bayesian algorithm. Remember that you are not the only email sender; billions of email messages are examined by these algorithms every second, and many of them share with filters used by other ISPs. Getting past a Bayesian filter can be done, but it just may be that your particular message may look like some other sender's spam, and, depending on what you need to communicate, yours may never go through. (Never is too final a word in the world of spam filters. We should say, never until something tells or trains the filter to see the message differently.)
- Never send email using a Gmail, Yahoo, AOL, or other "throw-away" domain. Use only a professional, branded domain, even if you have to create a new one for your campaign. It is wise to use a different company to register your domain, because if you do manage to get it blocklisted (it happens), this may insulate your business from harm.
- It's best to use a 3rd party SMTP (In other words, send bulk emails on a different sending platform, such as through an email sending service). Also, many marketers use a FROM email address with a different domain than the main business domain. With the combined presence of your business traffic plus your bulk email traffic, you are more likely to trigger anti-spam systems.
- Even if you are using a domain that has been established for years, it is important to "warm up" the domain. That is, send only 15 to 30 emails at first, gradually building up to 100 to 200 per day. This will help prevent email servers from thinking that you've been hacked and a spammer is pushing out huge spam loads on your domain. A service like lemwarm* can help with the warming process.
- If you do get your domain blocklisted by a particular ISP or email server, even your business email may end up in the junk mail folder. Worse, if your domain gets blacklisted, you may not be able to get ordinary business email delivered at all until the blacklist is cleared, which can be a long and expensive process.
There are plenty of other "tricks" that seasoned emailers swear by, and plenty of them are probably useful, but marketers rarely have the time to learn and employ them all. Sending email campaigns means getting bounces. A bounce rate of 30% does not mean that 30% of the email addresses were invalid or of poor quality; it means only that they didn't get through. Unless they return the result "Invalid," you may consider them worth trying again with a different message. If you get an "Invalid" rate under 10%, your send was good, even if your bounce rate was significantly higher.
A Word About "Open Rate" and New Email Issues
Updated July 22,2021
Up until recently, the "open rate" - the fraction of bulk email that has been opened by addressees - was the gold standard for judging the effectiveness of an email campaign. This is rapidly becoming a problem, because more and more ISPs and email service (like Apple Mail) are removing the components of the email message that are used to track opens. For instance, with iOS 15 arriving this Fall, the Apple Mail app* will start running all images through a proxy server to remove the tiny image-based tracking pixels that report when and where messages were opened. 45% of Apple owners, and 50% of healthcare providers, use Apple Mail, so this is a big deal.
Furthermore, Apple Mail will also hide user IP addresses, which is used to determine locations. Traffic will run through Apple servers and a third-party server to remove identifying information. Traffic sent from the user's device will be encrypted so that third parties can't see what users are searching for (and that supposedly includes Apple too). Apple users who select the "Protect my inbox" in the native Apple email app will change the way their email is download, in particular, prefetch with download all email content, thus making it appear that your email message has been opened. This will impact the "open rate" measure to an extent that will render it nearly meaningless. The email headers will all be stripped, removing our ability to view user information. (Clicks, i.e., tracking links, won't be affected.)
Read the article, "Apple's 'Mail Privacy Protection' is an Earthquake for Email Marketers"...
Another iOS 15 feature, dubbed "Hide My Email," will allow users to obfuscate their true email address when filling out forms on our websites, which will result in a loss of collected email addresses as users turn off the "fake" email addresses whne they start receiving too much email on those addresses. Note that about 80% of new healthcare provider mobile app environmentss are iOS, which presents the native email app by default, and experience has shown that most users retain this. Over time, we might expect these considerations to reduce email communications between marketers and HCPs to be clamped down significantly. This will have the effect of reducing the "noise" in the email channel, which can be valuable to marketers long-term, in being able to reach our most appreciative audiences without disappearing in a fog of abusing emailers. When our content and engagement are strong, email ROI will continue to be high.
Apple will not be alone in making these changes, and, taken together, they will render the "open rate" much less meaningful. Moving forward, the "click" is the new "open." And because the filters generate machine clicks to check content for bad actors, you can't really trust the first click; when you get a second click within an hour or two, it's that second click that means an actual human saw your message.
These comments, by the way, are based on the beta stage of the new iOS 15. The actual release in mid-September is likely to be different in at least a few ways, and CarePrecise will update this information as we learn more.
* CarePrecise does not endorse or advise against particular companies or services, nor does it receive payments for mention of any product or service.