I use user interviews for user profiling and assumptions verifications. Profiling goes through generic questions (who are you ? how old ? what's your job ? and so on...), and specific ones (how would you use this feature ? what device are you using ? and so on...). Assumptions verifications target product specifities (how is this feature working ? do you have any problems regarding this product ? and so on...). 

The one thing to avoid during interviews is causing biases. In order to do so, I exclude leading questions (DON'T : would you say that this feature is OK ? DO : how is this feature working ?). For statistics purpose, I also pay attention to ask every interviewee the exact same questions. 

Once I collect interviews data, I can turn them into several deliverables : personae, user journeys, global statistics, or detailed presentations. 

I identify two ways to use these data :

• the discardable way is time efficient but doesn't allow global data review as iit is only used for a specific period of time
• the atomic-research way is time-consuming but allows data synergy and global utilization