Could testing operate separately from other members of process or how to involve internal customers in QA process?

In my presentation I will focus once again on the issue that high- quality software can not operate without involvement of all participants of quality assuarance. We get used to the fact that usually in testing could be involved a lot of participants, such as analysts, testers, sometimes business users or maybe all the abovementioned at the same time. But is it enough to ensure full and complete audit of software?