How are you improving your GHL conversation AI responses?

Running conversation AI on a few sub-accounts and the results are hit or miss. Some leads get handled fine, others get completely lost on basic questions.

I want to compare what the bot is saying versus what our team used to say before we turned it on. But going through each conversation manually inside GHL isn't realistic when you've got thousands of threads.

Anyone doing this? How are you pulling the conversation data out and what does your process look like?

3 comments