With Conversation Analytics, you can measure your team’s responsiveness to your customers’ messages on Interakt. As a business, after you have taken the wise decision of including WhatsApp as a crucial channel for customer communication, you need to start measuring and improving that communication as well. Customers expect super quick responses and resolution via WhatsApp, and if your business provides that, they will be willing to buy from you.
Definitions of Metrics:
Total Conversations: Number of conversations initiated by customers in the selected period.
What is a conversation?
A chat on Interakt is made up of different conversations.
Every time you close a chat (by clicking the tick mark at the top right of the chat panel), a conversation ends. The 1st message sent by you or your customer in a closed chat starts a new conversation again.
Note: We don’t consider notifications sent via campaigns or any automated messages (like out-of-office messages) as part of a conversation.
Understanding Conversation Analytics
The Conversation Analytics dashboard provides a comprehensive view of how your team interacts with customers. It helps you track response times, resolution rates, and the effectiveness of your automated messaging, ensuring your customers receive timely support.
Overview of Key Metrics
The Overview tab is divided into three primary sections to help you gauge overall performance at a glance:
1. Automation Messages Sent
This section tracks how often your automated workflows are triggered.
- Out of Office / Welcome / Delayed Messages: Monitor the volume of standard automated replies.
- Workflow Conversations: Tracks specific automated flows built for WhatsApp.
- Custom Auto Replies: Counts responses triggered by specific keywords or customer inputs.
2. Performance Stats
- Total Conversations: The total number of customer-initiated chats within the selected period.
- Responded: The percentage and number of chats that received a manual response from an agent.
- Resolved: The number of chats marked as “Closed.”
- Note: To get meaningful insights, ensure that your team members consistently close chats once an issue is resolved.
3. Response & Resolution Time
- Wait Time for 1st Agent Response: Visualizes how long customers wait for their very first human interaction.
- Average Wait Time: The typical duration between customer messages and agent replies throughout a conversation.
- Resolution Time: The total time taken from the first customer message to the moment the chat is closed.
Analyzing Agent Performance
Switch to the Agent Performance tab to see a granular breakdown of how individual team members are handling their assigned chats.
Metric | Description |
Assigned | Total chats directed to the agent. |
Responded | Number of assigned chats the agent actually replied to. |
Total Resolved | A combination of chats the agent closed or reassigned. |
1st Response Time | How quickly the agent sends the initial reply. |
How to Filter and Export Data
Use the control bar at the top right of the dashboard to customize your view:
- Events Filter: Narrow down data based on specific campaign events like “Reminder for Free Credits” or “Campaign Update.”
- Tags Filter: Filter conversations by specific tags, such as manual upload dates or customer categories.
- Date Range: Select a timeframe (e.g., Last 7 days, Last 30 days) to view historical trends.
- Export Data: Click this button to download a CSV file of your analytics for offline reporting or further analysis.
Best Practices for Accurate Data
- Close Your Chats: The “Resolved” metric relies entirely on agents manually closing conversations.
- Monitor ‘0’ Stats: If you see “0” in your Responded or Resolved boxes while Total Conversations are high, it may indicate a bottleneck in agent assignment or a need to re-train staff on closing tickets.
It is very important that your team actively closes chats when the chat does not require their attention anymore. Unless this is done, the analytics won’t be meaningful for you.
Why do we say so?
Suppose a customer reached out with an issue on 13 July at 2 PM that was solved on 13 July itself within 15 minutes (2 pm to 2:15 pm). Similarly, the resolution time for the same customer on two other dates – 15 July and 17 July was 10 minutes each. However, if the chat hadn’t been closed on those 3 days, the entire exchange would have still been 1 unresolved conversation, which in turn, wouldn’t be an accurate representation of reality.
Responded: Out of customer-initiated conversations, those conversations that were responded to, by your team, in the selected period.
Resolved: Number of customer-initiated conversations that were responded to and closed in the selected period.
Closed without Response: Number of customer-initiated conversations that were closed without a response in the selected period. Check out Example 2 below to see an instance of how this metric is calculated.
Wait time for first agent response: Time is taken to respond to the customer’s first message in a customer-initiated conversation. On the dashboard, we show a median of 1st response times across different conversations.
Average Wait time for agent responses: Average time taken to respond to all customer messages in a customer-initiated conversation. Then, a median of the averages obtained from different conversations is taken.
Note: Only closed conversations are considered for this metric.
Resolution time: Time taken to resolve customer-initiated conversations (duration between the 1st and last messages in the conversation). On the dashboard, we show a median of resolution times across different conversations.
Still confused? We have explained this below using 2 simple examples:
Example 1:
Total Conversations: 3
Responded: 3
Resolved: 3
Closed without Response: 0
1st Response Time:
=> Median of (2:10 – 2), (4:10 – 4), (3:20 – 3)
=> Median of (10, 10, 20)
=> 10 mins
Average Response Time:
=> Median of:
Average (2:10 – 2), (4:15 – 3)
Average (4:10 – 4)
Average (3:20 – 3)
=> Median of (42.5, 10, 20)
=> 20 mins
Resolution Time:
=> Median of (4:15 – 2), (4:10 – 4), (3:20 – 3)
=> Median of (135, 10, 20)
=> 20 mins
Example 2
Total Conversations: 2
Responded: 1
Resolved: 1
Closed without Response: 1
1st Response Time:
=> Median of (2:15 – 2)
=> 15 mins
Average Response Time:
=> Median of:
Average (2:15 – 2), (4:10 – 4)
=> Median of 12.5 mins
=> 12.5 mins
Resolution Time:
=> Median of (4:10 – 2)
=> 140 mins
If you want to see data for each conversation, you can click on ‘Export Data’ to obtain a CSV of all conversations in the selected time period. Here, you can conduct your own analysis to get a deeper understanding of how your customers are being responded to.
We strongly suggest setting targets for each of these metrics and aligning your team on those targets. We will soon be sending you weekly emails with a summary of the week’s performance. Having a weekly review with the team should be helpful in gradually achieving your set targets and enhancing your customer’s experience on your WhatsApp channel!
We hope this article helped you understand how the Conversation Analytics work on Interakt’ WhatsApp Business Platform.
