Make sure you are coding for meaning, not content: check your codes are not simply summaries of utterances.

Don’t get hung up on coding: it is a process to help make sense of the data, not the aim of analysis.

Constantly ask of your emerging hypotheses about ‘what is going on’: what would counter this? How can I further test this theory?

Don’t move too quickly to ‘recommendations’ or ‘relevance to health and wellbeing’: these will take care of themselves if you do a thorough analysis, and rushing to ‘implications’ can curtail a good analysis.