This post originally appeared on Deliverability.com. You can read original posts from this author, and plenty others, on their site once a month.
How should you be expected to manage an email program without an understanding of how it has been doing? The reality is you can’t. Without having a constant measurement of how your email programs are performing, it will be difficult to understand how the ever-changing landscape of your user base, content, and email filtering is impacting your revenue.
Let’s take a look at two of the most common and successful ways of measuring deliverability.
Analytics Across ISPs
Internet Service Providers (ISPs) don’t provide any metrics as to whether an email landed in the inbox or bulk/spam folder of the end recipient. Without a direct measurement, we need to look at indirect methods. The first we’ll discuss is looking at different measurements across the ISPs that you send to. Here’s an example:
As you can see, if you measure your open or click rates across ISPs over time, you’ll be able to quickly identify any anomalies in deliverability and take corrective action. This same method can be applied to any metric that correlates to your deliverability rates like opens, clicks or even revenue depending on your business model.
The challenge with this method is you’re often trying to solve a problem retroactively. By the time you have enough data to understand the impact to your deliverability, the reason might have been something that has already passed like poor content or higher than normal complaints.
Sample With Seeds
The next method is using ‘seed boxes’ – email addresses at various ISPs that are set up in an effort to understand deliverability. The thought is these seed boxes will act like your customer’s inboxes and if you send a small sample of an email campaign to these boxes which you can access, you’ll be able to measure your deliverability by logging in and seeing the message in the inbox or spam/bulk folder.
However, this method does come with challenges as well. It’s difficult to get enough boxes in order to make your sample set large enough to consider statistically significant and you’ll want to ensure your seed boxes are sterile enough to ensure they have not been contaminated by email engagement (clicking, reading a lot of messages, complaints, etc.)
As long as you understand the pros and cons of the different methods of measuring deliverability, you’ll be able to utilize the best blend of analytics for your email program. If there are other methods you’re using for measuring deliverability, drop me an email and you might just see them in a future post!
Mike Veilleux is the Director of Product Management for Dyn, a cloud-based Internet Performance company that helps companies monitor, control, and optimize online infrastructure for an exceptional end-user experience. Follow Mike on Twitter: @MikeVeilleux and @Dyn.