You first need to decide on the scope and purposes(s) of the review, and how you will define the central terms and concepts, as well as the use you will put the review to. You need to decide on the languages in which literature will be reviewed and also whether some journals outputs reflect reviewing processes that are too poor to allow articles to be included. You need to decide whether or not any of the quantitative analyses within the articles reviewed will be aggregated to produce a synthetic result. all this is the review protocol, which must be transparent to readers.

Then you need to choose the databases which you will use. Google Scholar is not going to cut it. Scopus and Web of Science will help. Besides using databases you will need to review key journals manually to assess how universal the database coverage is.

Overall, I recommend David Tranfield's summary of how to do it published some years ago in the British Journal of Management. The best reviews normally appear in Journal of Management and can be used as a model.

First of all, you should select a theoretical framework to help you restrict the focus. Then you should read the classics on your chosen topic, even if it's to discard them or deconstruct them later on. Then you look at the intersections between your chosen topic and the theoretical framework you have selected, and from there you start going through sources. Personally, I find that when you start you will soon find a multitude of sources and from them more sources. You should also check if somebody has already a literature review regarding your topic and contact expert in that field to ask for recommendations.

It would help to know what is the topic and what is the theoretical framework.

You first need to decide on the scope and purposes(s) of the review, and how you will define the central terms and concepts, as well as the use you will put the review to. You need to decide on the languages in which literature will be reviewed and also whether some journals outputs reflect reviewing processes that are too poor to allow articles to be included. You need to decide whether or not any of the quantitative analyses within the articles reviewed will be aggregated to produce a synthetic result. all this is the review protocol, which must be transparent to readers.

Then you need to choose the databases which you will use. Google Scholar is not going to cut it. Scopus and Web of Science will help. Besides using databases you will need to review key journals manually to assess how universal the database coverage is.

Overall, I recommend David Tranfield's summary of how to do it published some years ago in the British Journal of Management. The best reviews normally appear in Journal of Management and can be used as a model.

I would first define your question. Then do the necessary literature review with defining inclusion and exclusion criteria. Then data extraction and data analysis follows. Some useful links for conducting systematic review are

Article Systematic reviews: How they work and how to use them

Article How to do a systematic review

You can also do an umbrella review or overview of reviews ( = systematic review of systematic reviews) if there are already many in your field.

Dear colleague:

Systematic review has two levels:

1. Broad Reading (this step is quick but very important, to make sure that your research will be new and non-repetitive, help you to identify the theoretical and methodological frameworks of your research, this reading also will help to know the cognitive time of previous research, to make sure the value of your research ) . These are paving road to next level.

2- Focused reading: I agree with answers of our colleagues. Erica Terranova , Richard Croucher, Konstantinos C Fragkos

Initially you have to be firm about your research topic. Then identify the main concepts and variables that you want to link, and analyze. Then , you can look for relevant literature. I do not think that one has to confine himself/herself within the boundary of any theoretical perspectives.

We used the online software Covidence for a scoping review. You just have to import all the references you need to do your systematic review and follow the steps. User friendly and precise.

I find the PRISMA protocol guidance for systematic reviews pretty detailed and helpful, you can find helpful resources here http://prisma-statement.org/Protocols/ProtocolGuidance

Be focus and at home with Theoretical, Conceptual, and Empirical issues surrounding your variables of study. Mapping these three items with a view to critique and identify lacuna would lead you to a good review.