Hi @ANDREW AURILIO! Working on getting you some input on this. In the meantime, this recap of a webinar we ran a couple months ago on Smart Trackers and Initiative Boards might be helpful! 

Hello again @ANDREW AURILIO! Here are some tips from @Vik Ekambaram: 

1. Gong uses AI to find concepts and to that end, the more calls (data) we have the more accurate the smart tracker.
2. When building a smart tracker, go for narrower themes. So you might end up with more STs than a keyword tracker. Example, if you are trying to understand if 'churn' is coming up in conversations. You might break this up into individual churn reasons and build a ST for each. E.g. Financial churn (you are too expensive, your competitor is giving me. lower price), Budget churn (we are going through a tough time, our budgets are cut), Product churn (there are feature gaps, your product has too many bugs etc.). As you can see each of these is a different concept and you cannot lump them together into 1 ST (the mistake I see customers make as this is how they built Keyword trackers).
3. Going back to point #1, you need a minimum number of calls for each sub-concept for the AI to work. By going narrow, you are reducing the number of calls.
4. Applying filters on the tracker will reduce the number of calls it will bring back. So you want to determine if the filter is relevant to the concept or if you can apply it later on an initiative board. E.g. if I am looking for churn conversations, I want to filter my calls to a stage that is post-sale (as a lot of pre-sale calls will not have churn discussions and if they do, they are in the wrong context).