Email : Taking Email's Measure
The most important metrics you're overlookingJanuary 2013 By G.B. Heidarsson
This should not be confused with what is often referred to as deliverability. As the word implies, deliverability only addresses the initial ability of the email to be delivered. That is, the sender receives verification that the email was transmitted from its servers/computers and was not immediately rejected by the receiving system.
That message may, however, be completely or partially rejected after evaluation from the receiving system, or it may be delivered to trash or spam folders rather than to the inboxes to which it was intended to be delivered. Thus, good deliverability does not guarantee anyone will actually see the email you sent, but bad deliverability guarantees much or all of the mail will not get through.
Until a couple of years ago, we relied solely on seed-based inbox delivery analytics. At the outset, this was a comprehensive solution where third-party monitoring companies would sign up for a multitude of random email addresses with different ISPs. By adding these addresses to the overall mix of subscribers—in some or all email deployments—the companies could use automated technology to log into the inboxes and see if the email had arrived, therefore giving the marketer a basic metric of inbox delivery success.
The accuracy of this approach was questioned by some because there were usually not enough of these "seeds" to give a comprehensive statistical view of success, but this method is still widely used today.
However, since ISPs have been enabling inboxes with both automated and user-defined personalization capabilities, the accuracy of the seed-based approach has waned. Instead, inbox delivery has been championed to combat this issue, giving a true picture of the consumer group being monitored and what is actually happening to the emails once they are sent.
Monitoring inbox delivery and immediately reacting to any issues is mission-critical, because everything else rests on you getting this one right. A 20 percent drop in inbox delivery can easily equal a 20 percent drop in revenue from that campaign. (This supposes everything is equal, meaning all names on your list have the same activity level and responsiveness.)
Opens and Clicks, Only Time Will Tell
You also need to look at your email analytics over time. That's what the ISPs do when they evaluate you as a sender. But, more importantly, your overall monthly metrics will tell you about the health of your list and your ability to sustain success.
Open rates for specific campaigns only tell you that a particular subject line was engaging, which is not very actionable information. Further, while outrageously successful subject lines increase your open rates, they usually do not result in outrageous sales increases, and responses to them typically decline rapidly over time. Descriptive, not-overly hyped subject lines that match the offer in the email copy are therefore the best approach. You don't want your subscribers to feel disappointed when they open the email, as that can prevent the sale.
Gaining an understanding of how many subscribers have opened one or more emails in a given month or quarter tells you a lot about your relevance and subscriber engagement. In essence, the goal is to have as many subscribers as possible engaged enough over time to open some of your emails.
With this information, you can work to maximize revenues or responses from your most engaged subscribers, re-engage inactive subscribers, or, if that is not successful, reduce the cadence for that segment.
If worse comes to worst, you should suppress them so they don't damage your mailing reputation with the ISPs. Sometimes a series of messages, clearly stating in the subject line that you will stop sending them emails if they don't open, can determine if they are completely disengaged or if they can be saved.
Although click rates tend to have a better correlation to revenue than open rates, click rates on a campaign-by-campaign basis are not very valuable, either. Of course, the better the offer, the higher the click rate. And often, the higher the click rates, the higher the revenue. But you don't need analytics to tell you that—you knew that before you sent the email.
What you want to understand is if you have a portion of your subscribers who open your emails but don't click through to your website. In that case, in my translation of Philip Kotler's "AIDA" model, you seem to have succeeded in getting their attention and interest, but failed in creating enough desire to click through and take action. (Here's a definition of AIDA)
If you have this happening, you need to reconsider your offer and make sure that your subscribers understand that you are a respectable vendor.
On the other hand, if you are actually getting the clicks through to your website, but not getting the conversions, the offer may again be where the problem lies. However, you should also make sure that you are doing everything in your power to reduce the perceived risk of doing business with you. Do you state that recipients' personal information and credit card information are safe? Do you have accreditation from third-party organizations that verify the integrity of your business? This sometimes plays a bigger role than you think.
Continually monitoring your subscriber groups' activity levels over time and using that information to control the frequency and cadence of your email programs will help you with the mission-critical task of getting into your subscribers' inboxes, as well as maximizing the benefits of your email programs. Once you are in the inbox, use your analytics to learn more about your audience—where they are in the customer lifecycle or purchase funnel and what data you could use to move the needle even further. This will help you maximize the monetary value of your subscribers and keep your list healthy.
G.B. Heidarsson is senior vice president at New York-based marketing data compiler eDataSource. Reach him at firstname.lastname@example.org.