Conversation
A Conversation is any service engagement or event generated by the Conversation Platform and administered by the Virtual Agent; it can be a chat, a voice call or a Service Request.
Common Scenarios
Below are some examples of common scenarios for analyzing Conversation data.
Conversations By Channel
Scenario: As a data consumer/analyst, I want to see the breakdown of conversations, as a function of conversation count for all Expedia partner/entities, to understand the outreach of customers on our various partners in the past 'x' number of days/months.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Channel Type
- Measure: Conversation (Count)
- Filter By: Conversation Started Date and/or any desired filter
Conversations By Product
Scenario: As a data consumer/analyst, I want to know the conversations by product serviced by the Conversation Platform.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Product Name
- Measure: Conversation (Count)
- Filter By: Conversation Started Date and/or any desired filter
Conversation Platform Adoption By Channel
Scenario: As a data consumer/analyst, I want to see the breakdown of conversations, as a function of channel type for different Expedia entities/partners, to understand adoption of different channels.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Partner Name (pivoted), Channel Type
- Measure: Conversation (Count)
- Filter By: Conversation Started Date and Partner Name
Conversation Platform Adoption By Entry Point
Scenario: As a data consumer/analyst, I want to see the customer engagement on different entry points to understand the adoption rate of various entry points.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Entry Point (pivoted), Launch Point
- Measure: Conversation (Count) *Filter By: Conversation Started Date and/or Partner Names
Agent Conversations
Scenario: As data consumer/analyst, I want to know the percentage of conversations (chat/voice) where a human agent is requested so that I can gauge the effectiveness of the Virtual Agent and make necessary improvements.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Agent Requested (pivoted), Channel Type
- Measure: Conversation (Count)
- Filter By: Conversation Started Date and Channel Type (in Voice, Chat)
Conversation Self-Service Rate
Scenario: As data consumer/analyst, I want to compare how many conversations were agent-assisted versus self-serviced for each product.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Agent Assisted (pivoted) and Product Name
- Measure: Conversation (Count)
- Filter By: Conversation Started Date and/or Product Name
Customer Engagement
Scenario: As data consumer/analyst, I want to know the customer's engagement on the platform so that I can research root cause and recommend corrective actions.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Customer Engagement
- Measure: Conversation (Count)
- Filter By: Conversation Started Date and any desired filter
This will display different types of customer engagement:
- 'Ghost' = Text Conversation has no inbound messages.
- VA Only' = Conversation is assigned to a VA, but never assigned to a Queue for an Agent.
- 'Abandoned in Queue' = Conversation is assigned to a Queue, but is never assigned to an Agent.
- 'No Responses to Agent' = Text Conversation is assigned to an Agent, but Customer sends 0 messages after queued.
- 'Engaged with Agent' = Conversation is assigned to an Agent.
- 'Outbound' = Agent initiated the conversation."
Contained Conversations
Scenario: As a data consumer/analyst, I want to know the percentage of conversations that are contained and those that are not.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Contained Yes/No
- Measure: Conversation (Count)
- Filter By: Conversation Started Date and any desired filter
Deflected Conversations
Scenario: As a data consumer/analyst, I want to know the percentage of deflected conversations. These are conversations that was initiated with a VA but eventually a human agent is requested.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Deflected Yes/No
- Measure: Conversation (Count)
- Filter By: Conversation Started Date and any desired filter
Completed Surveys
Scenario: As a data consumer/analyst, I want to know the number of surveys completed from those that were offered in order to gauge effectiveness of our surveys.
How: To achieve this, select the following in the Customer Experience Explore V2:
- Dimension: Customer Engagement
- Measure: Survey Offered (Count), Survey Completed(Count)
- Filter By: Survey Offered Count is greater than zero, Conversation Started Date, and any desired filter
Data Dictionary
For a list of available fields and their descriptions, refer to the Conversation Data Dictionary.