Is it because of many filters, or because too many components, or because too heavy pql queries?

Because I think the number of components (and thus queries) is the main reason for slowness, at least in my experience.

Maybe you can check the load with the borwser console, maybe you can get some clue.

I think that moving all the filters to the Sheet or Analysis level shouldn't impact to the PQL query time, as I think it will compose all the filters into the same query (so, no matter if those filters come from Analysis, from Sheet, or from component, it will put them all together to the PQL query for that component) (not sure if I explained myself... )

HTH

Hi Guillermo,

thanks for the reply and your ideas. You really got my point and I see that only moving the filters to the analysis lvl would have very low impact so I will better save my time for some query optimization. Main reason are definiely heavy pql queries within filters + in KPIs. Often is used the FILTER_TO_NULL function to make the PUs filter aware and here I see the most heavy part...however it will be a challenge to change it