Understanding Average in Microsoft Social Listening

Microsoft Social Listening (MSL) is coming - and you can read more in my previous blog posts "Are you learning? Analyzing using Microsoft Social Listening" and "Are you listening?"

In the last few weeks I've gotten a few questions about the average curve(s) in MSL, so I'll explain.

In this example I'm monitoring the posts on the "AX 2012 R3 Launch" (note the upper left corner in the screenshot below) and looking at the volume history for the last week (note the upper right corner). The day I'm writing this article i May 13th, hence my chart is from Tuesday May 6th till Monday May 12th (="Last Week"). 

The red curve is the number of posts per day of the last week (25, 23, 14, 27, 1, 1, 6)

The chart also displays a grey curve for reference; the average curve (5, 5, 15, 31, 12, 2, 13). The average curve is always in terms of the current time frame, and calculated for the last five time frames. Since my current time frame is "Last Week" this grey curve is for the last five weeks.

So the first "5" on the grey curve is the average of the number of posts in the last five Tuesdays (May 6th, April 29th, 22nd, 15th, and 8th) - eg. (6+4+5+7+3) divided by 5 = 5

Reading the chart

So what does this chart tell me? The number of posts this past Tuesday (25) and Wednesday (23) was higher than the 'average' Tuesdays (5) and Wednesdays (5) in the last five weeks (left side of chart). Friday apparently is the 'peak' day (center of chart) of the week when it comes to number of posts regarding the "AX 2012 R3 Launch" whereas weekends is more 'silent' (right side of chart).

Exporting

In case you need to export the above chart (or any chart in Microsoft Social Listening) you can click the arrow in the upper right area of the chart

And your exported file (jpeg in my example) would look like this (ready for your PowerPoint presentations). Note, you can also export the numbers and greate your own charts)

I hope you'll enjoy MSL once its here :-)

Updated

  • Try out Social Listening - link (30 minutes session with real live environment)