Hi @diego m​ , it really depends on what kind of customer service you providing. 

1. Resolution time also matters;

2. Conversation rating

3. Busy hours, day's, month's 

4. Conversation per team member

5. Tag's usage statistic

Here some great blog posts about this topic:

https://www.intercom.com/blog/how-to-keep-kpis-meaningful-for-your-customer-support-team//p>

https://www.intercom.com/blog/5-support-metrics-that-matter//p>

Generally i prefer to look at the live versus offline experience and double down on that, so less focus on resolution time and more on;

1. Responsiveness: First Response Time (as an SLA) + Follow Up Response Time (each subsequent update)
2. Customer Satisfaction: Agent SAT

From there we then focus on metrics as Roy called out and blend stuff like;

• Agent availability (% of time available against rostered time)
• Mean Time to Resolve
• Conversations closed
• Conversations participated in
• Conversation complexity mix (conversation attributes captured define simple, hard, complex requests)
• Quality score

There is a bunch of out of the box historic reporting in Intercom, but alot of what we need is done manually outside of the system to ensure health of chat as a service.

@craig​ @roy s11​ 

thank you guys - valuable inputs!