By Brooke Chloe
Research has shown that businesses which make use of data driven decision making, big data and predictive analysis tend be competitive and have higher returns than their counterparts. This is why some of the biggest companies constantly want to obtain more data from consumers as well as employees. However monitoring employees and gathering information from their data can be a difficult and tricky task. So how can a company collect such data with respect to the usage of time, the activities which take place and the relationships at work while at the same time respecting the boundaries of their employees and their personal information? Here’s how:
1. Finding a sponsor: the team which proposes data analysis should have the motivation to change their business based on the findings which they receive. Thus, a number of businesses require a sponsor from a senior position to gain this kind of support. This person can help in balancing out opportunistic quick wins with a view of how their predictive analysis would fit into the strategic plan. They would also have to explain why the collection of data and its analysis is important for employees throughout the organization and would serve as the person responsible to ensure that the data remains private.
2. Come up with a hypothesis: prior to collecting data, decide why you are in need of it. Legal departments of any company are unlikely to approve a project if it does not have any objective. Apart from this, the team which proposes this project should have a clear understanding of what they wish to gain. This includes knowing why the data is need and what changes would be made once the findings come through. It would also consist of how the results of the enforced change would be measured as well as the return on the investment which would be put in that would justify the energy and time put into the project.
3. Anonymity and aggregation: the sender of the data and the recipients email addresses should be made anonymous with respect to their departments. In order to further protect the anonymity of these individuals, reporting should be aggregated into a minimum grouping size which would not make it possible to drill down to one person’s data in particular and no guesses would be made as to who they are. This would get rid of the possibility of even the slightest amount of snooping.
4. If employees cannot be anonymous, let them decide how the data is used: there are times when the objectives of the business do not allow data to be kept anonymous. Thus in such cases, the best method would be to ask for permission prior to gathering data. This can be done in two ways; send employees an email stating they will be included in the study and giving them the option to opt out of it or, giving them the opportunity to opt to take part in the study
5. Look out for information which may be confidential and then screen again: departments such as the legal, HR and those working on mergers would require the data to have a great deal of protection. Thus, information that is gather irrespective of its collection method should be screened in 2 ways; information should not be gathered by configuring the instrument to exclude characteristics, keywords or individuals which would suggest sensitivity or, remove any data which was not screened in the first configuration as both the software and the people can misinterpret the meaning behind the text. Prior to sharing data with whom it is intended for, conduct a second validation
6. Don’t look for personal information: irrespective of the values employed by your company, the personal lives of employees should not be looked into and such information should automatically be removed from the dataset. Employees are seen to be having a right to their privacy along with legal rights which differ from country to country. Analytics initiatives should not look into personal matters.
7. Make use of a third party: in order to prevent privacy violation of the employee, it is common to make use of a third party which is responsible for data cleansing, anonymity and aggregation.