One of our favourite books about research is Silverman's "little pink book" in which he engages in a discussion about "what is data?", "where do we find it?" and "what then do we do with it?"
In particular, he distinguishes between what he calls "naturally occurring data" and "manufactured data". "Manufactured data" is the most commonly collected in qualitative research: which includes interviews, focus groups, and observations. It is said to be "manufactured" because we, the researcher and the subject, together are creating, or manufacturing data in a particular setting. "Naturally occurring data" on the other hand, is data that is collected after is has naturally occurred: which includes social media data, policy documents, websites, newspaper articles etc.
We believe there is nothing "natural" about putting someone in an non-descript office to interview them with a red flashing light flickering at them on a little recording device!
We see that most researchers quickly choose to do interviews in their research design as a first point of call, often without considering other methods of collecting data. Interviews, and even more so focus groups, are the most difficult data collection methods for a variety of reasons (access, timings, resources required). Thus we find it puzzling that these are the first port of call in most qualitative studies.
We consider Silverman's advice, to first collect "naturally occurring data" where possible, as it is often more easily accessible (often publicly available), and then to support this with "manufactured data" that we generate through interactions with the subject.
Our advice: consider the pros and cons of each of the possible methods in your planning, which helps to decide which methods will yield the most useful data for your study. Doing this will help you to map out your research questions, with the most appropriate methods to answer them, which will help you to justify your choices (and non-choices).