Livechat Dashboard
Engagely.ai is a conversation AI Product that helps businesses make their chatbots, voice bots, and social media bots. It has an extended product called LiveChat. Using LiveChat, the business’s customers can connect directly to human agents in case the bot is not able to answer any questions.
About
My role
User Experience Design
Data Visualization
UI Design
Tools & Technologies
Adobe XD
Duration
2 months
Problem Area
The platform was plagued by long queue lines, inefficient query resolution, misallocation of agents to unfamiliar fields, agents taking breaks while customers wait, underutilization of chat resources, receiving queries outside of business hours, and customers experiencing dissatisfaction due to unfulfilled expectations.
How might we
reduce customer wait time, improve query resolution time, increase chat utilization efficiency, and minimize transferred chats and calls?
Solution
Gather data
Collaborate with data analysts on the different types of data that was received on the back end.
Design & iterate
Work on lo-fi wireframes and visual designs to show data in the best possible way for maximum insights
Measuring success & impact
Based on the goals set in the initial stages, we measured success for the 2nd phase by comparing it with data from the MVP 1 phase
Visual Design
Metrics were divided into two parts: Real-time (kept updating every minute) and Historical.
Real-time dashboard
Aims to reduce wait time when a customer waits for a time limit above the specified time (option to customize this on the portal). Sends a notification to the manager if an agent was not responding within the time limit. Aids manager to go to the chat and take actions like assigning chats to someone else, reply to the chat, etc.
Historical data - Response trends
Average First response time, Response time and Resolution time which could be filtered by time. team, agent and channel
Historical data - Agent Performance
The manager is also able to see each agent’s performance for the time period like the assigned chats, avg response, resolution, wait time of the agent, no of chats assigned, efficiency, the CSAT score given by the customer to the agent, etc
Player game screen
Question with answers written in white cards are given to players to select from. A timer to reduce the wait time is given to all the players.
Leaderboard
Leaderboard can be seen by dragging the bottom layer and the leaderboard can be seen easily with the person having the highest points at the top.
Historical data - Agent Activity
The agent’s activity like the time online, offline, on break and idle was shown per day
Queued Customers
The agent’s activity like the time online, offline, on break and idle was shown per day
For the agent
Another part of the solution was for the agent. We tried to map the sentiments of the customers who were talking to the agent as well as the sentiments of the agent who replied to the customers. Angry, happy, sad, and neutral were the basic sentiments. This helped the agent understand how to react to a particular customer. This also helped the manager analyze if the agents are responding appropriately.
Impact
30%
Reduction in resolution time
24%
Average session time reduction
80%
CSAT score above 4/5