Currently I am sending most of my campaigns to my whole list. Generally performance (open and click rate) is fine but I am looking to divide my list in an engaged segment and a reignite segment. This way the engaged segment should bring more or less the same absolute performance as the whole list (relative performance will be a lot higher). And the reignite segment has the objective to bring people back to the engaged segment or else these people will go to the suppressed list.
My question is, is there a function in Klaviyo to find the best variables and timeframe to segment my engaged list on. So can you run some kind of regression to find dependent variables that have a significant impact on for example opens.
I would think that with all available data you could find that for example: ‘x opens in the last y days’ and ‘x times active on website in the last y days’ has the biggest chance that a user will open the next email.
Thank in advance!
Best answer by alex.hongView original
Welcome to the Community.
At present there wouldn’t be a way to segment your audiences automatically based on engagement. However, despite not being able to segment based on a premade algorithm, you can come close to seeing how engaged contacts are based on the number of emails they have received in a given time frame and how many emails were opened. From there, you should be able to manually calculate the percentage out.
For example, if you had a segment definition of “What someone has done, received email equals 10 times in the last 30 days AND opened at least 7 times in the last 30 days” this would mean the contacts caught by this segment would have an open rate of 70% or greater in the last 30 days. Although this is not a true accurate representation of one’s engagement level and open rate, it will come close to achieving your goal of segmentation by percentages.
I would further recommend taking a look at some Help Center articles Klaviyo offers on the subject of segmenting for engagement. I’ve included some below for your convenience:
Hope that helped,