Why it matters and What to try

Rules decide what gets flagged. AI only writes the explanation (Why it matters) and the fixes (What to try), using that flag's own numbers and best practices from the ad platforms' documentation and published research.

Before an explanation is shown, PipeHalo checks that every number in it matches the flag's data and that nothing in it is unsupported. If it doesn't pass, the flag shows its plain sentence and standard fixes instead.

Each explanation says when it was written. If one looks wrong, click Something wrong with this explanation? and tell us. Your data is never used to train AI models.