Here is the general principle for citations: anything that is not considered "elementary knowledge" to the community reading the paper must have a citation.
This means that different communities need citations for different things.
For example, when I write for a biology community, I often end up including citations to basic computer science textbooks, to give a grounding for things that are simple undergraduate information in my background, but that would be highly unintuitive to an experimental biologist. The reverse applies as well, for a biologist writing to a computer science audience.
Something that you know through experience definitely needs to be backed up with citation. One of the critical values of the scientific method is that it lets us separate statements supportable by facts from myths based on cognitive biases. We humans are very, very bad at learning from experience, in the sense that we draw many conclusions that are simply not true. Folklore is full of these types of experience-grounded myths, some of which turn out to be true (willow bark does help with pain: that's the basis for aspirin), and some of which turn out to be false (mercury turns out to be terrible as a medicine).
If you can find things backing up your experience in reliable literature, you can cite them. If not, perhaps you have a good subject for study, if you can figure out how to design an appropriate experiment...