- we're automatically adding a "country" tag to each new conversation with a country code (for example "ca" for canada)
- We have a department per language (so english department, french, italian etc)
- in some markets we have a bot . in those markets - the conversation is first routed to the bot department (English bot) and when needed - the bot handsoff the conversation to the "English" department and the human agents take over.
- the bot was built using the "conversations api" of zendesk, and is treated (it seems) like a normal human agent
- some of our reports are produced by the GoodData dashboard within zendesk
- Many of the main metrics we use to analyse our agents / country performance are now skewed as the bot is always the first agent. for example "chat reply time", first response time, response time etc
is there a way to have a bot live, but still get the REAL performance of a department (or any other grouping of conversations like tags) ?
i know that there are many companies using bots so im assuming this was solved already and would love to hear how! :-) thanks!
Please sign in to leave a comment.