I am posting this as more of a discussion post, rather than a question, as I am trying to come up with ideas for a project I have been asked to consider.
For context, I work for a customer experience (complaints etc) for a large retail company, and we are dicussing new ways to measure and understand our customer experience week on week. This is to be fed into the wider business to offer context to our performance etc.
Currently, we choose our "important topics" based on volume that week (other things come into play but for ease of this post, I shall stick with volume). Every week we see the same 3 topics of complaint hit the top three "biggest volume" spot.
We measure this against performance YoY, MoM, WoW etc, but the stories are stale. We know for certain that every week complaint volumes for delivery delays (for example) will be in that top 3. But do we care? We know it's going to be there, so is it interesting enough?
My ask here is for ideas, nuggets of knowledge or theories to go away and look into.
As way of example, let's say this week we have the same old top 3 complaint drivers... But there is a fourth complaint topic hidden in all our data that has suddenly appeared! The volumes are low, but the variance has rocketed WoW. How would I measure the impact or "interesting-ness" of this topic?
Granted I can just look at the variances of the topics and eyeball anything that looks interesting, but is there a better way? A more data driven way to identify these topics?
I have access to a plethora of data from within the business, so I am really open to anything you can throw at me.
Once again, I appriciate this is more of a discussion topic than a question per se. so I appricate any help or guidance you may be able to offer.