Out of scale value

Question

Hello,Have you ever experienced an outlier in the line chat that makes you impossible to read the normal values? I'm referring to this chart

Jennifer_Dungan · Accepted Answer

Yes.

I am assuming that was an invalid spike? And not something like a one day flash sale or something that went viral causing a real spike?

Sometimes what I will do, is to create an ad hoc segment to identify fraudulent spike data and exclude it within the report. I try to limit this to only the day/time where the spike occurred (using the Day / Hour / Hour of Day / etc dimensions to "time block" it). I will make a text note saying that the exclusion rule may not be 100% accurate, but to it was done to the best of my ability to "right size" the data.

Imbalakumar · Answer

Hi @leonardopo

Yes, I do experiences when the data is deviated lot from the normal trend. In these cases, exclude that deviated data(percent in this case), then you will align the trend like the below

Its normal while we get tracking or calculated metrics like this.

Thanks,

Bala

Sign up

Login with SSO

Login to the community

Login with SSO