There isn’t always a one-size-fits-all when it comes to reports. For this reason, we have a setting that allows you to choose between Average, Median, or Percentile. Average is the most well-known, but it can be easily skewed by outliers. Our new Median option helps with the outliers, and Percentile allows you to dig in even further to see your median for a certain percentage of your customers.

Average (mean)

Previously, the average (in maths terms called the mean) was used, which simply adds up all the numbers and divides it by how many numbers were in the dataset. This works great for datasets where all numbers are similar, but if it contains one or more outliers (i.e. numbers that are extremely high or extremely low compared to the rest) it will drive the average either upward (with high outliers) or downward (with low outliers).

Recommended for teams where there is consistent flow of incoming conversations, and you want the average of all conversations.

Median

The median fixes this by sorting the numbers by size and taking the value that is in the middle. That way, an outlier won’t skew the average number, and you’ll have a better view of what the majority is.

Recommended for teams where there may be some conversations that are outliers from the norm, and you don’t want those conversations to skew the metrics.

Percentile

Lastly, the percentile works similar to the median, except that you can get the number at any point in the dataset except just from the middle. For example, this allows you to set it at 80 to see what the metrics are like for 80% of your customers.

Recommended for teams where you want to see what the metrics look like for a certain percentage of your customers.

 

 

It’s an organization-level setting to ensure everyone is looking at the same data, and it can be customized under Settings → More → Reporting.