Help shape “Extremely Negative” signal sensitivity and events tuning

​@bradybluhm 

How should we balance sensitivity and control so these signals are trusted, actionable, and aligned to your org’s workflows?

Supporting questions:

1. If we offered presets (loose/medium/strict), how would you want to apply them: globally, by persona (exec/manager/CSM), by account tier, or another way?
   • I could honestly see use cases for all of the above.  For example, depending on your definition of loose/medium/strict, an exec might only need strict, where as maybe a named CSM would want loose. Alternatively, by account tier, maybe strict for your long tail customers and loose for your enterprise level customers.  Maybe also by topic?  So, loose for xyz topic, and strict for abc topic?
2. What feedback controls are most valuable: submitting “should trigger/should not trigger” examples, renaming event types, consolidating similar events, or routing rules (who sees what and where)?
   • I think initially “should/should not”, and then, consolidating similar.
3. In your day-to-day use, what minimal context do you need with each event (e.g., key quote, who said it, artifact link, severity rationale) to decide whether to act without opening the full thread?
   • Minimal: key quote and artifact link with the ability to “drill through” to view severity rationale

IMHO - 

1. need some weighting to the source - ie a Support case should have less because that already has a path to escalation that doesnt involve CS 
   1. could consider excluding Support from extremely negative 
2. Need to consider customer persona weighting as well 
   1. ie Exec sponsor negative messages carry far more weight than a user support case 
3. Alignment of negative message to lifecycle.. 
   1. ie a negative message in a renewal conversation when customer is in renewal lifelcycle should be flagged 
4. Feedback should be available within
   1. default reports cards that can be added/removed from persona based dashboards
   2. auto -create tasks (similar to AI follow up in CS)  for Account CSM to review and close out or convert to CTAs 
5. Would like to have ability to train model in a more scalable manner than by individual event
   1. Reporting on root causes similar to Topics Report would be really useful 

Regards

Shaun

How should we balance sensitivity and control so these signals are trusted, actionable, and aligned to your org’s workflows?

Supporting questions:

1. If we offered presets (loose/medium/strict), how would you want to apply them: globally, by persona (exec/manager/CSM), by account tier, or another way?
   1. This is difficult to answer well because we’d likely leverage any/all of what you provided. As ​@heather_hansen pointed out, in the case of an Executive, we’d use strict - like “only things they need to see/be involved in”. And tier is a fantastic suggestion - we’re already leveraging different settings for different tiers in our Account Dark/No Reach Out insights, so this would just mesh perfectly.
2. What feedback controls are most valuable: submitting “should trigger/should not trigger” examples, renaming event types, consolidating similar events, or routing rules (who sees what and where)?
   1. I would prefer “should trigger/should not trigger”, but would definitely use “consolidating similar events” if it were available. 
3. In your day-to-day use, what minimal context do you need with each event (e.g., key quote, who said it, artifact link, severity rationale) to decide whether to act without opening the full thread?
   1. Absolutely require the key quote and perhaps the stakeholder title (this would help in determining how to act, for instance, if the CEO said something versus the Technical Administrator, we’d be able to determine different approaches for each stakeholder). And, then ideally something that gave us the “why” Staircase thinks this is impactful. 

Also, I could just repeat everything ​@smartin said - because he’s spot on. I am also going to re-express my interest in admin controlled naming of “extremely negative messages”. Though, I would still want to have something called “extremely negative messages,” I just want something that’s a little less Chicken Little - “possible friction, please look at me!” kind of thing.