I designed a flow to ask user input values through a series of questions.
Halfway through, should the user disrupts the flow and wants to clarify something else - how should we design the chatbot to handle such “ad-hoc” / “out of the flow” cases?
Disrupting the flow
I find that it’s almost impossible to do that in one flow! It is possible but it’d become so complex that only you (as someone who knows what’s happening behind the scenes) can go through the clarification and jump back in and continue the questionnaire.
This is mainly because Flowxo doesn’t allow the Trigger Flow action to do deep-linking between flows, meaning that I could use the Flow action to trigger a Label half-way through another flow. I personally don’t see a need for this because the problem you’re facing comes down to a UX design question and how good your on-boarding is before asking the user to give you information.
You need to explain yourself and the bot well enough that the user doesn’t feel the need to clarify anything halfway through answering questions.
Of course, there will be outliers no matter how good you design your pre-questionnaire segment, but it all comes down to reducing the proportion of those outliers. It’s better to have a bot that breaks for 2% of people than a bot that breaks for 20%
My suggestion is to user test a ton of different on-boarding situations and see how many of your users diverge half-way through the questionnaire after each on-boarding scenario. After all, bot sucess is mostly a test of how compelling your micro-copy is.