Bot + Human Conversations

Hi, guys!

We’re building a bot for Messenger and we need to enable a human to take control of the conversation if the user ask for it. We’re thinking in two ways to call a human: by saying something like “talk to human” and an option in the menu.

How are you guys dealing with such interaction?

Thanks a lot! :slight_smile:

Hi, yip that’s exactly how we are doing it :slight_smile:

Would love to hear if anyone else has other ideas.

Both those options are valid, but you should also consider identifying when the bot gets stuck or doesn’t understand, and proactively does something (either suggesting something to the user, or notifying a human to intervene).

(We have this built into our platform so you can build a custom workflow depending on what happens).

1 Like

I agree with Paul & Tony, having a global option that can be called either through text or a menu makes sure that the user can always get additional support. Suggesting that a human be looped in when the bot ‘fails’ is also recommended; I would add here that you qualify when the ‘human’ suggestion should enter the picture, e.g. After two attempts (rather than one, to provide an opportunity to gather extra data on the cases that the bot doesn’t understand).

1 Like

Take control when the Bot doesn’t understand should be your main way… I’m developing a same service bot, I’m alerting to the dashboard agents when the bot doesn’t understand and the bot is freeze meanwhile. Soon I will public my service in