1

I have two sets of monthly data that spans over almost 2 years. The subscriber volume, and the customer complaint volume. I'm trying to determine the best way to adjust the changes in complaints volume for the growth in subscribers, so I can compare YoY growth in complaints, without the impact that more subscribers would have on the absolute volumes.

Previously I had just calculated complaints/subscriber. But I'm wondering if there's a more accurate way that would account for subscriber growth.

I've tried applying the same technique as you would to calculate Real GDP using the subscriber volume as the base, but I'm not sure the method applies the same way. I've spent a while looking online for different ways to approach but many don't really apply or are too complicated for me.

I'm just using excel, and my statistical knowledge is limited to a few classes (understand regression etc, but nothing too complicated).

kjetil b halvorsen
  • 77,844
garfield
  • 11
  • 1

2 Answers2

0

Previously I had just calculated complaints/subscriber.

This seems like the right way to do it. If you actually get more or less complaints per subscriber as your number of subscribers increases, you probably don't want to normalize this away.

Kodiologist
  • 20,116
0

Another approach: Use some generalized linear model with log link function (maybe gamma, if complaints are counts maybe Poisson or negative binomial), the most important detail is the log link. For details see Goodness of fit and which model to choose linear regression or Poisson.

Then use the $\log\left( \text{number of subscribers} \right)$ as an offset.

If this is too complicated just go with the answer of @Kodiologist.

kjetil b halvorsen
  • 77,844