Coming back to this after having some time to think on it. I have no real data to go with my theory, but looking at it from a processing aspect. It would seem that: 1. "assign to AD groups directly, the agent queries AD first and then retrieves the actions matched with current users/groups to execute" would be able to match faster compared to "assign the action to everyone and use filter/condition, the actions will be retrieved always and the agent will skip or execute it after the filter is evaluated." 2. Just a theory though, but I did see yesterday when a customer has a good bit of filters/Conditions for AD groups ( match this or match that, or match both, it can take up to 20 seconds or longer to go through and apply what it needs to do on each log in for NP machines. Granted that the WEM cache is stored on the C drive, and my mind continues to go back to persistent location and the use case that "Use accelerate for cache" would help here more, vs keeping it on the C drive. 3. I keep going back to this. "_Use Cache to Accelerate Actions Processing_ This option can significantly reduce logon times, as the agents no longer need to connect to infrastructure services to retrieve relevant settings. Instead, agents can use the settings in the local cache, resulting in faster processing of actions and a faster logon time" Well, if you keep it on the C drive for NP device, there is nothing to accelerate action processing, because the cache is being pulled down again each login. The article goes into stating "By specifying a persistent location, the “LocalAgentCache.db” file will not be recovered after restarting the machine, and the WEM agent doesn’t need to perform a full synchronization every time the machine starts up. Additionally, choosing a persistent location ensures that local optimization data and statistical data stored in LocalAgentDatabase.db will not be lost after restarting the agent" I know WEM will still Function if you don't do this, but will it speed up these actions, and the WEM GPO processing that many of us are seeing that is hit and miss on applying at the right time? To clarify, even more and ask in a more direct way will the GPO settings apply faster if the cache persist? This seems to be another area of delays myself and many other have noticed. The article give a PVS example, and it can be taken that hey that only applies to PVS, but I think they just gave an example in the use case. Again I love WEM, But I am trying to really put logic behind how to make it as quick as possible when it's heavy used in environments that pertains to NP machines only. I also understand that I could be really overthinking this. But ideally it's really around the most optimal setup for speed around settings. I had a customer a couple weeks ago, where the GPOs were not applying 100% again from a WEM side compared to the GPMC they were previously using, and they were disgruntle about it and let me know how they felt. For me it's hard to hear this, because in the back of my head, I seen this many times in other environments. So my thought is, well is it they way the Cache disk is being handled here? Or is there a timing issue going on and maybe people are just not using the GPO side because of the native MS GPMC just works for the computer side? IDK, just my thoughts and looking for some tips here. Maybe @Anthony Shi or @Yan Wang https://www.citrix.com/blogs/2023/07/11/improve-environment-stability-and-performance-with-wem-agent-caching/

Found it. Thanks man. I missed this.