Beyond the basics of delivery and engagement, many email marketers don’t dig deeper to better understand what their metrics are really made of. As a result, many wins and losses are hidden in the aggregate metrics. Within an overall disappointing clickthrough rate, for instance, there may be impressively high clickthrough rates for one or two segments.
Indexing, a method of calculating the individual success of a specific subgroup within a larger database, can help you recognize and act upon information that may be lost in the cumulative data.
The easiest way to present the concept of indexing as it relates to email marketing is deliverability. Many email service providers now segment their deliverability reporting by ISP. Often, though, this reporting does not provide feedback on inbox delivery, so a marketer is often stuck assuming he is doing well across the board because the available metrics don’t indicate otherwise. In this case, reviewing the index of a subgroup’s open rate is a great tool for diagnosing potential issues that may be hiding in the aggregate.
To index for inbox delivery:
1) Assume that all subscriber activity is distributed equally among the ISPs. In other words, assume that a specific percentage of subscribers, regardless of their ISP, will never open an email, will sometimes open an email, will always open an email, and so on.
2) Find the distribution for an ISP across the entire database. Determine the percentage of valid and subscribed addresses that are @hotmail.com, @yahoo.com, @gmail.com, and so on. Let’s call this metric 1.
3) Since we’re looking to measure inbox delivery, we should evaluate engagement as a proxy. Therefore, calculate opens by ISP as a percentage of all opens—that is, what percentage of all opens are @hotmail.com, @yahoo.com, @gmail.com, etc. This is our metric 2.
4) Now divide metric 2 by metric 1 for each ISP. The result is the index of opens by ISP. (You can expand this to include clickthroughs as well.)
Looking at the example in the chart at the bottom of the page, the Sympatico domain is underperforming, as it should generate approximately 11.8% of all opens, rather than 8.2%. On the other hand, subscribers at the Videotron domain open 40% more than expected. With these metrics in hand, a marketer can begin work on resolving potential causes of the underperformance and thus increase the overall success of its program. For example, are Sympatico emails being delivered to the inbox? Do the emails render correctly for Sympatico recipients?
You can repeat this exercise for any number of measures. Perhaps most useful is reviewing engagement by the source of the email subscription: existing customers, sign-up forms on the Website, coregistration, contest entries.
With marketing dollars more focused on return on investment, it’s important to identify low-performing segments. By examining how subgroups within the larger database are performing, you can invest more in those segments where the returns are greatest and withdraw resources from that that perform poorly.
George Adamidis is principal at Real Email Consulting.
