This is community wiki question.

I will be writing my first review for ZBMath. I would like to take some suggestion through examples.

In general, abstract is too small and introduction is too lengthy in a paper to decide whether to read the paper or not.

I am under the impression that an ideal review for a (published) article should be more like an invitation from the referee to the user to read the paper, or, much better to be an appreciation letter.

Question : Are there any reviews in ZBMath (Mathscinet has restricted access) that motivated you to read the article. Can you give some examples.

I am aware of the question How to write a good MathSciNet review?. That discusses theoretical aspect of a good review but it not all of the answers have examples showing a good review.