In the past couple of months, we at Loyalty Lab have been called in to rescue some pretty sketchy e-mail lists. Our clients had been operating on a “don’t ask; don’t learn” basis for some time, clinging to the false confidence that comes from not having any way to measure the deliverability and performance of their e-mail marketing campaigns.
It was only when they migrated to an e-mail provider that was able to give them explicit data on deliverability and performance that they realized that their time living in a world of peaceful naiveté had come to an end.
Now let’s be fair. In none of these cases had the list owners intentionally done anything wrong, deceptive, or dishonest. But, without the knowledge and resources to constantly maintain a list that is free of spam traps, clean of abandoned accounts and populated with a low percentage of complainers, e-mail marketing lists inevitably will slip into the depths of low deliverability and poor performance.
Getting to know the companies revealed a number of commonalities that boded poorly for their e-mail marketing campaigns. First, while acquisition practices all had recently shifted to rigorous, permission-based tactics, each had a legacy of adding names and e-mail addresses using pretty creative means. A critical lesson for e-mail marketers is what is okay in the world of catalog direct marketing is not okay in our CAN SPAM driven e-mail world. For catalogers who have been renting lists and trading names among demographically similar merchants, there is no e-mail analogy. E-mail addresses must carry clear and traceable permission to market. While marketers will differ in the rigor of single and double opt-ins, there is zero room for non opt-in.
A second commonality was the obvious presence of a number of “spam trap” e-mail addresses. Addresses containing “spam” or “trap” often are used by ISPs to identify senders who have been harvesting e-mail addresses – at the worst from unaware chat rooms, more often than not from lack of attention to list hygiene. E-mail addresses of all sorts find their way into large lists. Attentive e-mail marketers regularly scan their lists for problematic addresses like these and purge them.
Finally, without a process to record and query which addresses are bouncing back or repeatedly non-responsive, the ISPs automatically suspect that something is fishy and start raising the deliverability bar.
Not long ago, ISPs filtered e-mail and made deliverability decisions (for example whether to allow into inboxes, spam folders, or simply reject) based on a complex set of content-related attributes. Did the e-mail have a subject line that promised a future of larger body parts, or a slimmer waist? Did the body contain tricks to deceive the filters like white on white html or extracts from Doctor Zhivago? Was there simply too much bold print?
While some content analysis still is used, the vast majority of e-mail deliverability filtering occurs based on a sender’s reputation. Reputation is built and destroyed based on a small number of factors. Most directly associated is whether or not receivers of a list owner’s e-mail complain to their ISPs. This is most commonly done by “marking as spam” or moving from the inbox to the spam or bulk folder. A number of ISPs have taken a secondary step to randomly ask their account holders whether or not certain e-mails were welcome or intrusive, ie. is this spam?
Next, reputation is affected by the cleanliness of lists. Mailing to abandoned e-mail boxes more than likely means a list owner has not had the capacity to log hard bounce replies from ISPs and therefore has not been able to take action to remove the names. But spammers continue to blast ISPs with millions of addresses looking for the small percentage that actually find their way in. ISPs have no way to know if you are blasting your way in or just lack the tools to keep your list clean. Spam traps are obvious sources of poor reputation: If your list has them, you will be caught.
What’s more, there is a lot of sharing among the ISPs and the services that they use to track and rate senders’ reputations. Just because you have not raised the eyebrows at one ISP doesn’t mean that your performance at another won’t eventually make its way back to ISP.
Saving a list’s reputation, and in doing so more successfully delivering e-mail, is a problem that we now regularly face. Simply changing your service provider to another IP range won’t help. Reputation is tracked back to the sender, not the servers from which the e-mail emanates. The good news is that we see dramatic improvement immediately and steady improvement over time.
Here is our recommended path to a better reputation and better inbox delivery.
- Purge your list of any e-mail addresses that were not acquired in strict accordance to CAN SPAM’s language. If you are in doubt, error on the side of safety.
- Perform a secondary scan and delete any names that could be construed as spam traps.
- Stop mailing to any name that has not open or clicked in the past six months. If you have not been tracking opens and clicks, start now and come back to this step in six months.
Since the three greatest factors of reputation are complaint rate, spam trapping and hard bounces, taking these steps will dramatically improve each of these factors and almost immediately positively impact your scoring.
To illustrate the impact, following these steps for one of our lists showed:
- Spam trap hits dropped from 12 to three
- Hard bounces dropped by 75%
- nbox delivery moved in the right direction at four of six problematic ISPs
The bottom line is, with the amount of effort and resource dedicated to building e-mail marketing lists, you must be equally diligent in maintaining your lists or you risk the very likely outcome of not being able to speak to your customers. The solutions are straightforward, a bit painful to swallow all at once, but clearly worth it in the long run.
David Rosen is senior vice president of Loyalty Lab (www.loyaltylab.com), a San Francisco-based developer of customer loyalty programs for the retail industry.