In Q3 2009 Symantec and ValueNetworks.com began using value network business intelligence approaches with the intent of improving the web-based customer experience and reducing the need for agent-based support. The Value Network Insights application was used to evaluate and process data from Symantec's customer support applications.
As a logical extension of ongoing efforts to synchronize agent- and web-based customer support, the primary goal of this project was to better understand the factors influencing customer usage of agent- and web-based support.
Another goal was to identify which KPIs or KPI combinations that are important to monitor. Further goals of the project were to determine:
What KPI patterns and thresholds are relevant for managing the customer support experience?
How might a tool for monitoring KPIs of the customer interaction interface look?
What might a tool for visualizing behavior of the customer interaction interface look like?
How should an alert and notification system be configured based on the above insights?
Results achieved:
Identified and validated network-centric combinations of KPIs relevant to monitoring and predicting the customer support experience. Currently these new indicators are expected to (semi-automatically) warn of problematic cases up to 90% earlier than the existing escalation system.
Defined a process for quickly extracting, analyzing, and visualizing data from the workflow system.
Designed a report to automatically highlight cases predicted to fail the SLA, to prioritize these, and to suggest escalation approaches for avoiding the failure. Reports can be created in under 10 minutes and are provided as a spreadsheet and a radar chart.
Results achieved by:
Configuring a default query to draw case data from the primary workflow application
Completing a statistical analysis of the Standard Value Network and Collaborative Value Network indicators, and then correlating them with operational performance measures (SLA met or not met)
Identifying the combination of KPIs, patterns, and thresholds best suited for both understanding why cases failed to meet SLAs and predicting which specific cases would fail to meet their SLA in the future
Establishing a Prediction Report creation process that provided repeatable results
Validating the findings of the Prediction Report with stakeholders
Defining enhancements to the Value Network Insights™ Enterprise Edition application to accelerate and support the report generation process
| 
