Collaboration activity patterns - measurement issue

AngieS
New Contributor
New Contributor

Hi community,

In the Wellbeing, balance and flexibility dashboard there is a graph displaying Collaboration activity patterns, we noticed some irregularities in terms of dashboards measurements, with the data not being aligned with experience of team analyzed and signals suggesting cases of much higher hours of activity and lower hours of inactivity. Has anyone else using this dashboard had any similar experience or got any idea of what could be causing the data in the graphs to be skewed and/or explanations related to measurements of why it seems not to reflecting reality?

1 ACCEPTED SOLUTION

Stacey_Fang
Viva Insights PM
Viva Insights PM

Hi AngieS,

Thank you for reaching out.

I'm not sure what exactly you are seeing in that chart, but one possible explanation is that active hour block (y axis) is calculated based on a week-aggregate approach. 

The report is run on weekly queries with one of them called hourly collaboration query where it shows employees' collaboration activities by channel (email, meeting, call, chat) by hour (9-10, 10-11 etc.). Since it's a week-aggregate report, as long as the employee had an activity during that hour block any day of the week (e.g., sent an email at 6-7AM on Wednesday and didn't collaborate during this hour rest of the week), that hour block will be counted as active for that week. This could potentially overcount people's active blocks which may result in what you saw as higher hours of activity and lower hours of inactivity. This was a trade-off made to avoid pulling significantly larger set of data which could impact the report performance. 

That said, we are working on a new solution to update this report and bring in a more accurate measurement of employees' active hours. Thanks for your understanding, and please stay tuned.

----------------------

If you'd like to explore how to measure it in a more accurate way in a custom way, please see below paragraph:

There is a way to measure the active blocks more accurately in a manual way. To go that route, you could run hourly collaboration query by day, add a column for each hour block to indicate whether the user is active during that block each day (e.g., if there's any activity - email, meeting, call, chat - during that block for the day, that block is active), add a column that counts the active block for the day. Then take the daily average by week. Map this daily average by week information with rest of the weekly queries for the report.

 

View solution in original post

1 REPLY 1

Stacey_Fang
Viva Insights PM
Viva Insights PM

Hi AngieS,

Thank you for reaching out.

I'm not sure what exactly you are seeing in that chart, but one possible explanation is that active hour block (y axis) is calculated based on a week-aggregate approach. 

The report is run on weekly queries with one of them called hourly collaboration query where it shows employees' collaboration activities by channel (email, meeting, call, chat) by hour (9-10, 10-11 etc.). Since it's a week-aggregate report, as long as the employee had an activity during that hour block any day of the week (e.g., sent an email at 6-7AM on Wednesday and didn't collaborate during this hour rest of the week), that hour block will be counted as active for that week. This could potentially overcount people's active blocks which may result in what you saw as higher hours of activity and lower hours of inactivity. This was a trade-off made to avoid pulling significantly larger set of data which could impact the report performance. 

That said, we are working on a new solution to update this report and bring in a more accurate measurement of employees' active hours. Thanks for your understanding, and please stay tuned.

----------------------

If you'd like to explore how to measure it in a more accurate way in a custom way, please see below paragraph:

There is a way to measure the active blocks more accurately in a manual way. To go that route, you could run hourly collaboration query by day, add a column for each hour block to indicate whether the user is active during that block each day (e.g., if there's any activity - email, meeting, call, chat - during that block for the day, that block is active), add a column that counts the active block for the day. Then take the daily average by week. Map this daily average by week information with rest of the weekly queries for the report.

 

Recommendations for you