Small description of output

README.md

@@ -8,4 +8,10 @@ Tokenizer and parser are made in a way to be used independantly and not tied
 with the output.
 
 Tokenizer use multibyte string function from PHP so it should support unicode
-without any issue.
+without any issue.
+
+## Output to docx
+
+Output to docx use custom styles such as `md-block-h1`, `md-block-h2`,
+`md-block-quote`, ...so you start with a a template document with all styles
+defined within and have your output match the visual design you want.