OVERVIEW
Background
Sprinklr offers customer service software that helps brands manage agents and customer interactions, to ultimately improve customer experience.
In this project, I created a simple and dynamic module for Automatic Call Distribution, which optimizes the flow for supervisors to manage agents, so that customer calls can be taken by available agents in the shortest possible time. As the sole designer, I worked closely with the product team and engineer team to make sure technical constraints, business goals, as well as visual impacts are in a good balance.
Impact
📉 The product has reduced contact center costs by up to 30%.
⏰ Agents are now able to serve customers 50% faster, with an average
response time decreased by 80% - from 5 hr 42 min to 70 min.
👏 Users from many enterprise companies have given good feedback on it.
Learn more: https://www.sprinklr.com/features/automatic-call-distribution/
PROBLEM
Agent resources are not being utilized to their fullest capacity.
Besides, data indicated that a lot of businesses reported
-
High Customer Abandonment Rate
-
Long Customer Wait Time
Why Does This Happen?
Temporary mismatch between case assignment and available agents:
Queues DO NOT get a static number of incoming calls at all time. In some cases like peak traffic or special occasions, one queue might get more customers calling in than usual at one time, which can result in lacking agents to answer. While agents responsible for queue A run out of capacity, agents responsible for queue B might be underloaded and free of calls, which led to uneven distribution of agents during certain day/time.
Identify Area of Improvement
Automatic Call Distribution (ACD) is a system that supervisors would use to set up algorithms for incoming calls and ultimately manage predefined settings of agents on calls, such as who are responsible for certain topics, and when they should be on call.
Once queues are set up, the ACD takes all of this information into consideration, scans it against the rules that have been configured, and makes decision about where the call should be directed.
Below is the flow of setting up a queue.
Meet the User
The Supervisor of a group of customer support call agents, who is in charge of:
-
Assigning agents to different queues
-
Tracking each queue's condition
-
Managing agents between queues
GOAL
Develop a more intelligent and adaptable distribution system for supervisors
EXPLORATIONS
Affordances
After design workshops, we reached to a decision on keeping the balance between users needs and dev constraints through allowing users to manually edit the queue.
Leverage with Existing Flow?
What if we incorporate this feature into the existing flow when the user creates the queue?
⛔️ No, because the use case is for temporary effort thus it should not mess up with the original setting.
Wireframes
SOLUTION
Enhanced Data Visualization
Colors are introduced in the table to indicate how occupied each queues is REAL-TIME
so that users can find out which ones need reallocation more easily.
Reallocation Capability
Users can reallocate agents from free queues to busy queues and set up the quantity and time for it.
Flexible and Editable Rules
Users can view and edit reallocation details through the primary table and the third pane.
The design of the third pane reconsiders the hierarchy of various information and highlights key information.
