All Collections
How do we detect Outliers?
How do we detect Outliers?

Here are the factors considered when detecting Outliers

Shruti Ramanujam avatar
Written by Shruti Ramanujam
Updated over a week ago

Determining if a piece of content is an Outlier:

We start by collecting all the posts from a brand in the past 12 weeks. Then we get the 90 percentile value of the metric we’re tracking like Likes, Comments, Retweets etc and then we multiply the value by a factor of 2. 

The result will be the threshold for that metric, therefore if a post has a metric with a value above the threshold, it will be considered an outlier. 

The Outliers you first see are for the past 7 days - we show 15 Outliers in total. Scroll down and click ‘Load more’ to see them all. After you see the 15 latest Outliers, we will show you the Outliers from the previous 7 days and so on. We update the list every day with more Outliers - the best performing Outlier will be found at the top of the page. 

Did this answer your question?