Catch All

This is where you set up the flow for instances when Instinct AI fails to understand the user's utterance without hampering the user experience. Chatman provides or is made with five reply types to deal with such situations:

Here is the explanation along with a scenario where the particular catch-all feature can be helpful. Choose the one that best suit customer needs and wants.

You can set only one Catch-All reply to all the intent in either of the response type. Choose that best suit your business and the customer needs and wants.

Human Handoff

Hand over the particular inbox to a human agent for better handling and improving customer experience. It's not always the case that the bot can answer every question.

Scenario

Say you're in the IT industry. Every time a user asks a complex question, the AI fails to understand. So you can set handoff as your catch-all so that as soon as the bot fails, the conversation is handed over to an agent or group.

How can you set catch-all to journey reply.

Select the option of Human Handoff.

Draft a message to be sent.

Choose the agent you want to handle this query.

Since you can set anyone catch all reply, it's advisable to use randomly so that any agent can be assigned, and if/she is unable to handle the query, he can be assigned to another agent.

The query can be assigned to a particular agent or in a queue or to a group of agents.

Reply

When Instinct AI fails to understand, a text reply involves providing a simple text-based response acknowledging the issue and asking the user to rephrase or clarify their intent.

Scenario

A user inputs a query with a different language or jargon that is not within the training. The Instinct AI responds with a friendly message, stating it couldn't fully understand the question, and asks the user to provide more details or rephrase the query for better assistance.

Select the option of Reply.

Draft a perfect message.

Add random variation, which will be picked up randomly by the AI to answer user utterances.

Journey

The bot responds with a particular Journey on the basis of query asked, thinking he/she might be looking for.

Scenario

Say you're into IT business. Many a time, users ask for setting up your software, so you can simply create a journey for that and set it as a catch-all. So whenever the bot fails, it'll respond back with the journey.

Include a hand-off card in the journey you create, in order to keep customer experience for all those whose intent wasn't to set up software.

Select Journey as an option.

Select the relevant journey from the drop-down list.

If you don't have a relevant one, create new journey.

insert

No Reply

In cases where Instinct AI is unable to understand the user's input and there's no suitable predefined response, the 'No Reply' option involves not providing any immediate response. This can be useful to avoid generating irrelevant or confusing messages.

Scenario

Imagine you're into finance; you expect the customer to ask a query with precise information so that the system can respond accordingly.

You can set this so that whenever a customer doesn't provide relevant and sufficient info, the system remain silent.

insert

Blocks

Blocks involve using predefined content or structured elements to guide the conversation. This can be of three types: list messages, reply messages and flow messages.

Scenario

Let's say you're in automobiles; whenever the Instinct AI fails to understand the user intent, you might want the system to respond with a regular message that showcases your services.

For this, you might set the reply block message with a body message and a CTA to your website or quick replies offering your services.

insert

PreviousIntents NextSummary

Last updated 1 year ago