It depends. If you are looking for quantitative, non-subjective criteria to rank journals, impact factor is a commonly used metric because it is easy to measure (albeit the case of journals falsifying their impact factor is not unheard of, so it may be necessary to use an independent source for this criterion instead of taking at face value whatever the journal says).
If you are evaluating journals in order to decide where to submit your next papers, impact factor may be an important criterion if you already know that this criterion ranks high in the evaluation of the quality of your papers at the institution where you work.
If you are not particularly concerned with how your institution evaluates the quality of your papers, the impact factor of the journal is just one of many factors. For example, submitting your MS to a journal well-read by scientists working in the same field (and therefore likely to give a high impact factor to your paper specifically) may be more important than the raw impact factor of the journal as a whole.
In Lithuania, not only IF - also quartile is very important. And quartile depends on the aggregated impact factor for the field - not only IF of the journal. Thus, journal in Forestry may have lower IF but be in higher quartile, than, say, journal in Ecology.
I know some unexpected rises of the journal to Q1 and then fall back next year. It is crazy world - when you send your manuscript, you have no idea, if it will be printed and when. however, best journals are almost "forever" :)
Yup, IF gives an idea about the quality of the journal. Inspite of yearly change of IF but usually it is a slight difference that didnt affect the quality of the journal
in the Clarivative (former Thomson's) you may check citations - are there many citations from the same journal. If there is, journal is rising it's IF artificially. Some Serbian journals were doing this, now these are not in Thomson;s