Kanban is a Japanese word for “signboards”. Toyota conceptualized Kanban and was also the first company to implement it. Kanban, simply put, is a work scheduling technique that helps Toyota sustain daily production targets without violating the Work in process inventory thresholds.
Possible Uptake of Kanban in IT Service Desk Management
There are lot of similarities between Manufacturing operations and IT Service desk operations. The flow of incidents through the service desk is analogous to the flow of material through various stages in the manufacturing process. A service desk can be seen as composed of multiple work centres manned with agents. Kanban can help service desk managers manage workload, locate and address bottlenecks, and reset service level expectations. By far, highest contribution to MTTR (Mean time to Repair) variability is incident investigation time.
The effect of various tactics to increase the rate of investigation can be monitored by adopting Kanban.
Track Incident Turnover
Kanban is also symbolic of other production control techniques that can benefit IT Service management. For example, one such metric is the Incident Turnover; built on the lines of inventory turnover, an inventory management metric. Incident Turnover is ratio of throughput and average backlog corresponding to a time period. Throughput is the total number of incidents closed in a time period. Sum of period opening and closing backlog divided by two gives the average backlog for that period. The ratio can be plotted using Quadrants (see image below) for each incident category. Service desk managers can employ tactics to improve throughput and reduce backlog to improve turnover of incident categories that require attention.
Let’s Move the Cheese!
The point of intersection between Throughput and Average Backlog can be seen as a cheese that is moved incrementally to form the basis of continuous improvement. The first point of intersection can be drawn based on heuristics related to individual categories. Actual turnover is most likely to fall in the third quadrant followed by the fourth and the second quadrant. Improvement in service operations can push the actual turnover towards the first quadrant. High incident turnover is a result of optimal utilization of resources, focus on automation and a utilitarian approach to knowledgebase management.
Backlog reduction without capacity expansion can be achieved through increased service automation. Diversity of incidents within high turnover categories should be investigated to identify scope for service automation. High volume of similar incidents are good candidates for service automation.
Capacity expansion could increase throughput but at a cost. Problem management could reduce recurring incidents.
Throughputs can go down due to various reasons like agent turnover, application or technology rollouts, service desk limitations, service desk downtime etc., Agent productivity can be increased with training, knowledge base, tools (chat / remote access).
Low throughputs and backlogs without service desk downtime and agent turnover either means the IT systems are maintenance free, robust, scalable and reliable or the business users have lost trust in the service operations. This dichotomy can be explained by launching customer satisfaction surveys.
Furthermore, Incident categories and assignment groups within the service organization can be stack ranked by incident turnover to discover any unexplained anomalies. Same analysis can be extended to more than one service organizations leading to development of industry benchmarks.
[Image credit: Wikimedia Commons.]