I have to start with a few questions such as do you recall your statistics class where we don't need to measure everyone and everything they did to get an answer within a degree of confidence. If you must know EXACTLY the percent with no margin of error then you have frankly gone off the deep end and must intrude on everyone's machine to install a "Mr. Burns" type app to report home with what you want.
Most folk will do the old Nielson rating method and pick out a few random candidates and use that for the sample set.
So my last question is going to inflame the non-statistician. Is this a Mr. Burns' method or are statistics good enough?
Currently, I am working on a time tracking project. I would like to see if you have any insights as to how you would approach this problem.
Current Situation: Global Management wants to see how employees at different regions (in different countries) are allocating their time without being invasive. The type of work these employees are doing are process/operational work, not consultancy/project type work. At the end of the month, management wants to see for example, at region Z, what % of worker's time is allocated for activity A and % of time is allocated for activity B. Or are region Z's workers more focused on activity A than region X's workers?
Any insight would be great. If you have any questions, do ask! Thank you.