In the previous entry, we covered the basic fallacies of using open rate as a primary method to judge the success/failure of an email campaign. As noted, a single method simply isn’t sufficient given the complexities of email marketing today. So, with that in mind, let’s now talk about how it should work.
External Mailbox Monitoring
Any good true statistical analysis begins with a control measurement. In the case of email, this is the seed email address. Seed addresses give you the ability to better gauge the overall success/failure of your delivery by checking across a variety of major ISP domains like aol.com, hotmail.com and yahoo.com. Companies like ReturnPath offer their own seed test list, which is constantly being update to ensure that your delivery tests are as accurate as possible.
In addition, email service providers, like us at mobileStorm, can do these seed address tests for you, both isolated to just a test campaign as well as placing the addresses scattered throughout your database for real-world email delivery checks.
Reconciling Opens with Delivery Rates
It seems a simple concept that if your opens are down that means your delivery must be down too right? Well, it’s not quite that cut and dry. This is why it’s a good idea to combine the external mailbox monitoring mentioned above with your open rates. If, for example, you see that your opens are going down gradually but your seed address tests are staying the same, this means that there could be an issue with the speed of your campaign delivery. In that case, ask your email service provider to reduce the speed or throttle of your delivery rate.
However, a much more effective delivery gauge is to check the ratio between your unique opens and the unique clicks. Those two figures should be relatively close to each other and if you see this stat start to deteriorate over time, it could be a delivery issue. This can be true even if your seed tests are still showing good delivery because until you do a true send, it’s still just an estimate in terms of your overall possible performance. Basically, when done this way, the numbers won’t lie. If you see decreases, it’s time to talk to your ESP about your delivery rates.
Another thing that affects your overall email campaign success/fail rates are the use of ISP feedback loops. Almost all of the major ISPs have implemented this system, which allows senders (both ESP as well as corporate in-house systems) to collect addresses that need to be removed from their database. This is important for a couple of reasons. First, it allows subscribers to use a system they are probably more familiar with – the ”this is spam” button or similar option on their ISP of choice – to remove themselves from an email vs. the footer. Second, and more importantly, it allows senders to ensure that people who are complaining to their ISPs to be removed from their database. This is critical because the ISPs various complaint level ratios to determine whether or not your email will get through or be blocked as spam. This, mind you, is just a very quick overview of feedback loops. In a future article, we’ll explain all of this in much more detail.
Dig Deeper in the Error Codes
For you techies out there, a final method to help determine your overall success rates with your email campaigns or lack thereof is to look at the raw error codes being generated. If you are using your own in-house solution, you can do this yourself by accessing the raw logs generated by the bounce file for a particular campaign. However, you have to be patient as wading through the codes can be very tedious. Nevertheless, it will give you increased insight into any potential issues. For example, AOL is very good at providing detailed explanations to any bounce that is generated, most notably spam bounces. This information points to various URLs which explain the error codes in detail. Again, if you have to do the work yourself, this information can be invaluable.
For those with an ESP, the raw logs are even more indispensable. Any ESP worth their salt will have access to them and use them as a key way to determine potential blocks/issues that could reduce your overall email campaign effectiveness. If, for example, you are seeing consistent issues with AOL, ask your ESP to check the raw bounce logs to see if there are any patterns to them in terms of the error codes. They should be able to tell you what the codes are indicating and more importantly, solutions to fix the issue. If not, then I would suggest looking for another provider.
As you can see, open rates only give you just a piece of the puzzle when trying to determine your email campaign results. In the final part of this series, we’ll address the message content itself – what can cause your success rates to drop and why, and how to avoid doing this in the future.
Until next time”¦