If you've purchased the software from the PRIMER company, it comes with a very well explanatory manual. You need to know hoe to import your data into the software (you can either import a distance matrix that you've obtained from other packages (such as UniFrac distances, Bray-Curtis, etc.) or you can import your OTU table and create matrix within PRIMER software (which won't allow you perform phylogenetic analysis such as UniFrac). Once you imported the data, then you need to define factors (that are different confounding variables that are specific to your experiment), You can find user-friendly tutorials in youtube that show how to import data and factors into PRIMER. Once you have imported your data and factors, you need to create a PERMANOVA model, which would be pretty similar to the ANOVA models that you have designed your experiment based on (you have the options to define random and fixed factors, as well as nested factors). In the PERMANOVA manual, there is a section that defines different experimental designs and how to create models for each of them.

Keith McGuinness from CDU has a Youtube playlist with a series of videos on how to analyse amplicon sequencing data and it's nicely explained (see link)

https://www.youtube.com/playlist?list=PL-W1dvsmEZ3KISwdig6X18dkQppJfcrYW

https://www.researchgate.net/post/What_are_the_criteria_for_selecting_variables_for_PCA_analysis_in_Primer_6_how_to_screen_variables