Secondary data in the form of : blogs, reviews, posts in the social media networks and print media, provide considerable data with least invasion. Next method: attending the company's events and functions and informal interaction or over-hearing of small talk. The data generated using the second method can not be validated. However, it gives some idea about the pulse of employees, their perceptions, attitudes and their current and future outlook
When thinking about non-invasive methods of gathering subjective quantitative data, I think a lot of recent ICT development are interesting to look at. Especially online questionnaires are becoming much more popular. For example, researchers can send employees emails with a link to an online survey. Although this is still invasive in terms of "we are doing a study and we need you to participate", employees can click the link when they feel like it and/or when they have the time to do so. This makes it a lot less invasive compared to the traditional paper-and-pencil test. However, there are some risks as well, particularly with regard to response rates as it is much easier to ignore these types of questionnaires.
I agree with Sridhar Sambatur for suggesting secondary data as least invasive method of collecting qualitative or quantitative employee data. In addition, collecting primary qualitative or quantitative employee data is not invasive as long as ethical consideration in research and confidentiality are observed and with clear and honest purpose/object of research. Any research method is invasive when it is morally and ethically injurious.
Especially intranet surveying is very interesting, however, not all workers have access to intranet. As such, one should care about who is in and who will be out of the survey.
I would like to add that intranet surveying, although it being a very promising way to gather data, really limits your research to workers with certain jobs and educational levels. A lot of workers with lower educational levels and/or lower skill levels, do work that does not require computers. As such, they often do not have work-related e-mail addressen or intranet at all, which means you could never do research among these groups of employees. Intranet surveying is therefore only suited when your research group would have constant acces to intranet at work.