Hello,
I was hoping someone could inform how the Avg ACW is calculated as shown on the "Agent Performance Detail" screen on Purecloud dashboard? Although the average acw converted times are matching, for whatever reason the data I am getting is slightly skewed by a few milliseconds and I believe the culprit might have to do with rounding.
Now the issue I am seeing which makes data comparison hard is the data is already calculated, averaged, and rounded once on the APD screen. But the data I am using will be the raw untouched data.
My understanding is Avg ACW = (total ACW / interactions with ACW). Which using PostAnalyticsConversationAggregateQuery and getting Tacw metric is all the data I will need. It should just be as simple as...(ACW Callback Sum + ACW Voice Sum) / (ACW Callback Count + ACW Voice Count). That is just basic averaging everyone learns.
This example above is what our BA put together per the data he is seeing on the dashboard by reverse engineering the calculated math.
Agent A
ACW average (As Shown on Agent Performance Detail)
Voice: 4:39 (20 calls) (Total ACW is 1:33:02, which is 4:39 * 20)
Callback: 0:05 (1 call)
Math:
Option A (using Avg ACW) ((4:39 * 20) + (0:05 * 1)) / (20 + 1)
Option B (using total ACW): (1:33:02 + 0:05 / 21)
Overall ACW Average 4:26 min
Time converted to milliseconds = 266,047.62.
Now my example based on the raw data.
Voice:
Sum - 5,581,857 ms
Count - 20
Callback:
Sum - 5,260 ms
Count - 1
Avg ACW - (5,581,857 + 5,260) / (21) = 266,053.1905 (rounded to 4th decimal)
Converted would be 4 Minutes 26 Seconds.
You can see my total in millisecond is about 5.5 milliseconds more. So this might not seem alot, but with hundred of employees, this only being 1 agent for a single day. When we go to do performance reviews, it can possibly skew the data by numerous minutes.
