As a starter, this work is often practically useful as a way of revealing what Ellen Isaacs from Xerox calls the 'hidden obvious' in her wonderful TEDxBroadway talk. From software development, to robotic surgery, to drivers-less cars, public sector organisations, small firms and massive corporations employ and engage ethnomethodologists to better understand the situated uses of technologies and the social problems and opportunities that arise from particular innovations. I work in a Business School and study work and the organisation of work in public and private sector firms. As a general point, EM/CA studies provide one way of 'bringing work back in' (Barley 2001) to the analysis of work organisations. Whenever you look at literature around an organisational problem or topic, it can be surprising how few people analyse 'what people actually do', the work people do, not in some general sense, but day to day. The vast majority of qualitative organisation studies is based upon what people say about what they do, and how they think about themselves and others. But, what people say about what they do, and what they actually do, are two different things. EM/CA studies bind concepts to 'work itself' (Strauss 1985).
Some EM/CA studies provide a 'respecification' (Garfinkel 1991) of 'academic' concepts. For example, when you look at literature on lying, you find its overwhelmingly done in experimental settings where, in contrast to ordinary life, 'truth' is unambiguous. In that literature, analysts develop academic methods to detect lies. But these aren't used by ordinary members of society. So, what do they use? How do people distinguish, not 'inquiries from greetings' or 'payments from donations', but 'truth from lies'. When you look, they don't use the techniques developed by academic researchers. By recovering the lay or 'ethno' methods people use to navigate particular concepts and methodological distinctions ('gifts' vs. 'commodities'; 'truth' vs. 'lies'), EM/CA work respecifies organisational, sociological and psychological concepts. The field of discursive psychology has entailed people taking this way of thinking and applying it to 'psychological' concepts. In consumer research, for example, consumer knowledge is overwhelmingly treated as a property of mind, rather than an embodied property of the way actions are composed. Re-specifying knowledge involves showing, as Heritage and others have done, how speakers give recipients access to what they know through the composition of their talk and embodied activity.
A lot of EM/CA studies are well positioned to contribute to literature on collaboration, embodiment, technology, co-operation, routines, teamwork, etc., because the data are almost always recordings of people working concertedly on tasks often using tools of some kind. Given the discussion of respecification above, it shouldn't be surprising to learn that many EM/CA studies have looked at classification work in general, and how organisational actors look at scenes and determine what an object is, and more generally what is happening. There has been huge efforts to really work on specific bits of interactional organisation, to show practices, and organisations of practice, through which interactional problems are resolved.
Finally, for some people, the work is intrinsically super interesting and rewarding. Take the Big Issue examples presented above, paraphrasing Harvey Sacks, if you don't find those interesting, goodness knows how uninteresting you'll find the rest of it! The business of at least trying to reveal the 'hidden obvious' is an interesting thing to do for a living.