Within the digital messaging industry, an opportunity will sometimes arise for an email marketing manager to send a re-activation campaign to dormant email addresses, in order to galvanize subscribers who don’t respond to your messaging.

From a deliverability perspective, some people feel that putting such a reactivation campaign on a shared IP address is OK. However, others feel that a dedicated IP is more effective, because it does not involve sharing IP space with your other, more reputable senders. In a shared environment, the cursory damage that might occur includes junking or blocking by ISPs due to old addresses that turn into recycled spam traps. Hitting a spam trap is not fun for any deliverability administrator and could have time-based ramifications for your more active senders.

To us, it’s not so much about shared vs dedicated, but more about volume and how long the inactive subscribers have been dormant. For example, if a client has a small number of inactive subscribers (say 100) that haven’t opened in 6+ months, then sending them a nicely crafted email saying, “this is our last communication” and throwing in an incentive to get them to engage should cause minimal fallout in a shared IP scenario.

Now, it’s an entirely different situation if you have 200,000 subscribers that haven’t had any activity in over a year. In that case, I would recommend taking a small control group and continuing to send to them as normal, probably on your dedicated IP. Remember that cadence matters to ISPs. Then, take a small random sample of the reactivation campaign and send an incentivized email. This email constitutes the true reactivation campaign to a segmented list. (Don’t forget to segment your inactives.) By doing this, you will find common denominators in your inactive pool that could shed light on the stillness. You may give each segment its own VirtualMTA for granularity. At that point, you can compare the control to the test group. If the test group does the same as the control, then you will have to decipher the benefits of moving forward.

Let’s assume that the subscribers you sent to in the test group have not shown any engagement. Then, in accordance with email etiquette, you would send them an “if-allowed”—one last email asking them to confirm their email to opt back in to your list. The “if-allowed” would be really geared toward the subscribers that haven’t engaged or opened or clicked in over a year, as six months is most likely not enough time to accurately gauge subscriber engagement.

The deliverability risks of old addresses that are turned into spam traps or syntax errors are just too great for anyone sending off of a shared pool. The easy way out is to send the reactivation campaign from a less reputable dedicated IP, but we don’t recommend this strategy either. It will probably do more harm than good.

If and when the time comes for you to reconfirm these addresses, to be safe, you could implement the process discussed earlier, and send them from a dedicated IP address, segmenting by domain and other parameters, and using VirtualMTAs for each segment. Think of the dedicated IP as an insurance policy, ensuring that you will not interrupt the flow of email from other good senders on your shared IP.

In these situations, there are no hard and fast rules. Email is a process and, after the send, listening to the stillness becomes your greatest asset. The goal is to come up with a smart plan for testing in order to pinpoint subscriber issues. Play with the segmentation; then, rinse and repeat. Staying proactive and always testing in your sending environment will help diminish the inactive roster. Stay patient and, ultimately, you will be able to see clear benefits from your campaign.