Unsubscribes: High-Level Analysis to Improve Marketing Strategy

[Tom_Kerlin2]

Hi all,

With Q4 underway, the marketing department has been asked to help with analyzing our unsubscribes (and unsubscribe rates) to see if there is any correlation with number of emails being sent. Specifically, we're trying to determine the number of emails an email address receives per day/week/month and the corresponding number of unsubscribes.

I wanted to know the most feasible way to go about collecting this data, whether its building smart lists or pulling reports (i.e. email performance report shows unsubscribe numbers and unsubscribe rates per asset).

Is it possible to pull data on number of times an email is sent to particular email address (i.e. lead) per day/week/month, or can this only be accomplished by going into the individual lead's Activity Log?

What are some unsubscribe trends that your organization has discovered?

Thanks!

Tom

[Josh_Hill13 · Accepted solution]

I did this exact thing over the summer. We did prove that sending more than 1 email per day increases the likelihood of an unsubscribe to that group about 20% vs. people who received 1 email that day, although we are talking about very small percents here. And the number of leads who received more than 1 email varied pretty widely each day.

We did this by

• Smart list of people sent email on a day at least 1x
• smart list of people sent email on a day at least 2x
• filtering out the cross overs since you can only use At Least, which is not the same as Once.
• running that against the # who unsubscribed from an email that day (and received an email that day)

So not super easy. Then you have to run the regression.

I suppose if you had Tableau, this could be easier.

I'd skip the analysis and just put on Communication Limits and consider removing leads who haven't engaged in the past 6 to 12 months.

[Tom_Kerlin2]

Thanks Josh.

So, would the smart list filter logic look something like this:

[Grégoire_Miche2]

Hi Tom,

This will give you the people who were sent at least 2 emails in a given period of time/.

-Greg

[Josh_Hill13]

yes.