Two Year’s Worth of Newsletter Tracking Data

In the comments from this month’s newsletter about managing a newsletter campaign, a reader asked,

Thanks for sharing all this data. Do you have a longer range of data for your newsletter tracking? I’m interested in speculating the reasons the numbers vary so widely.

So, I decided to put together all the tracking data from the past two years to see if any trends emerged among the relationships between the day of the week and the time of day on which we sent our newsletter out, and the amount of readers that actually clicked the ‘read more’ link in the email. In the graph above (click it to view the full size image), each month from the past two years is shown across the horizontal axis. Along the vertical, the day of the week is represented by the green lines, the time of day (AM from 12 down to the bottom, PM from 12 up to the top) by the blue lines, and the tracking numbers are in black and plotted along the black jagged line. The darker gray vertical bars are highlighting three particular points I want to look at…

The first thing that stands out is that our highest tracking number came in January, 2008 for our newsletter about Gmail. What’s really strange, though, is that we sent out this email on a Sunday at 4:30pm! That stands against the general wisdom, which says to send out your newsletters mid-week at around 10am.

You can see that our next highest point, in terms of the tracking numbers, came in January of 2009 for our newsletter on how to use Google Analytics. This time around, we followed that conventional wisdom and sent out the email on a Thursday at 10am.

Finally, it’s interesting to look at a recent low point too. Our lowest tracking numbers in the past year and a half came in October, 2008, for our newsletter on managing your online reputation. It seems pretty clear that sending out the email on a Friday at 8am was not a smart move. The fact that it was Halloween probably only compounded the mistake.

One other metric that the graph doesn’t show is the topic of the newsletter. I’ve wondered before if there might be a relationship between clickthrough rate and the topic (more appealing topics being read more, less appealing topics being read less). After looking at this, I’m still not sure if there’s a trend to identify, but some correspondences do make sense. For example, the most popular one shown was the one about Gmail. At the time, Gmail was still new to a lot of people, but getting a lot of press. Maybe many of our readers were interested in maybe starting to use Gmail. On the other hand, the topic of online reputation management was a bit on the fringe, so I wasn’t terribly surprised at the low numbers there, especially given when I sent it out. Lastly, Google Analytics was a red-hot topic, so I very much expected this newsletter to get a lot of interest. This is something I’ll be interested in keeping an eye on.