This is a suggestion of steps for a systematic literature review:
a) define the research subject, question, objectives, methodology and other general aspects;
b) define the academic databases that will be used for the search, such a Google Scholar;
c) define the queries (search words or expressions) that will be used on the search;
d) define (before and while the search is going on) inclusion and exclusion criteria (such as dates, areas, terms, age etc.)
e) run the first searches to determine if there are many or few returns and though refine the queries;
f) select or exclude texts based on reading the titles and abstracts;
g) read the texts to confirm selection and start writing about the texts;
h) categorize the texts;
i) analyze and interpret the texts;
j) write the review.
These are not rigid steps, that is to say, you can overlap some of these phases or even change its order, besides introducing other strategies, but they give a general sense of what a systematic literature review should look like.
It is also important to state that you can review literature in at least four aspects of your paper:
b) general theoretical references;
c) historical aspects of the subject (valid only in some types of research);
d) state of the art: people who are doing similar researches as your proposal.
Sometimes researchers believe that by doing a theoretical review, the other reviews are automatically done, but this is not true. It is really important that your paper includes these different types of literature review.