Botmake.io

Flows allow for short-term conversational linking, which adheres to the following format:

WHEN BOT SAYS: Hi
AND USER SAYS: Hi
THEN BOT SAYS: How are you?

Flows are designed to detect exact answers to Text 2, provided Text 2 is the answer to Text 1. If there is any variation on the user’s answer, the “chain” will become unlinked and the bot will depend on whatever default response has been set for the variation. So, if you want a bot that asks ‘How are you?’ after the guest user says ‘Hi’, you will need to account for all variations of ‘Hi’ and design a flow for each to competently cover the most common greetings.

The bright side to this, is that basic conversations can follow a principle of predictability, and with enough thinking, you can easily daisy-chain the conversation. For instance:

WHEN BOT SAYS: Hello
AND USER SAYS: Hello
THEN BOT SAYS: How are you? <- End of chain 1
WHEN BOT SAYS: How are you? <- Start of chain 2
AND USER SAYS: Good
THEN BOT SAYS: That’s good!

By copying the bot’s last line from the first flow and using it as the first line in the second flow, a basic “chain” has been created, which in this case distinguishes the user’s “good” as a contextual response to ‘How are you?’ as opposed to more a general use of the word. In theory, this means you can employ different uses of the word “good” based on context, provided you have accounted for other scenarios in which a Guest User might respond to your bot with the word “good”.

Here’s an example of account for multiple possibilities of the word “good” –

WHEN BOT SAYS: How are you?
AND USER SAYS: Good
THEN BOT SAYS: That’s good!

WHEN BOT SAYS: I promise I won’t tell anyone.
AND USER SAYS: Good
THEN BOT SAYS: …Not even a soul.

WHEN BOT SAYS: I’ll stop talking, then.
AND USER SAYS: Good
THEN BOT SAYS: <Zips lips>