How to write the state of the art without getting lost

The state of the art — or literature review — is the section where most students get stuck. Not for lack of sources, but for lack of structure. Here's a method to build it with order and a critical eye.

1. Start with a question, not a search

Before opening Google Scholar, define what you want to answer. A good review question filters the relevant and discards the noise. Searching "machine learning in sales" drowns you in results; "classification models for churn prediction in retail" finds what you need.

2. Organize by themes, not by author

The classic mistake is listing authors one after another ("Smith says…, Jones says…"). That structure argues nothing. Group sources by subtopic or approach and, within each group, compare positions.

3. Be critical, not descriptive

A good review doesn't just summarize — it evaluates. What approaches exist? What are their limitations? What gap does the literature leave that your thesis aims to fill? That gap justifies your research.

4. Close with a map and a gap

End by synthesizing what you reviewed and clearly pointing out the gap your work addresses. That transition connects the review to your problem statement.