Most of us have had interactions with service desks, and most of us can easily recall experiences that represent, in one way or another, fairly common frustrations. For example, when Rebecca needed a replacement laptop, she called up the service desk and made arrangements to have the replacement sent—quickly. It never came. She called again…and again…and again, each time speaking to different agents who confidently assured her it was on its way. But even as her frustration grew, she was quite sure there were frustrations on the service provider’s side as well. She could imagine the agents asking themselves why this customer kept calling about a standard request that the system showed as being in process. It takes time to investigate these kinds of exceptions (Did someone miss an approval? Was the mailing address incorrect?), and if their operation is like many we see, they could be working under enormous time pressures, perhaps with ticket backlogs, and finding the time to handle exceptions is a real problem.
Sometimes, no matter which side of the service desk we’re on, it feels like we spend a lot of time handling these types of issues. Is everything an exception to a standard process? Putting aside our own frustrations for a moment, we have to adopt a big-picture view, where we look at the entire end-to-end process like a workflow, where customer calls, tickets, and information flow through the service desk on an assembly line of sorts. Ideally, our work process should be standardized in such a way that these elements flow in an orderly fashion, although we know that even assembly lines can experience delays, backlogs, and bottlenecks in processing, as well as defects in the final product. Fortunately, there is a way to design workflows that standardize processes while allowing for a variety of outcomes.
The 3V Model: Volume, Variety, and Variation
Borrowing from manufacturing, we begin with a simple conceptual model that can help us understand the dynamics underlying many workflow problems. The 3V Model is derived from the discipline of operations management and it looks at operational workflows in terms of three different variables:
- The volume of customer requests (ticket or call volumes, representing the customer demand)
- The variety of customer requests (types of tickets or calls, representing the number of different customer requests, each one of these different types typically requiring a different way of handling)
- Process variation, or how the request is handled (the amount of variation in handling requests, reflecting the separate handling requirements associated with different ticket types, different ways of handling requests of the same type, and exception handling of one sort or another)
In an environment where tasks are highly repetitive, as we might see when there are large numbers of the same types of requests, we would expect the same types of tickets to be handled the same way each time. After all, to be efficient, high-volume tasks should have standardized processes. But very often this is not what we find, and the reasons underlying this process variation are sometimes surprisingly unplanned: people who do the same thing don’t communicate with each other, standardization was never a priority, employee turnover discourages continuity, no investment in centralized repositories for documented procedures, and so on.
Looking at problems in service desk operations through the 3V lens allows us to consider the influence of each in turn when we’re dealing with scenarios where the interactions between these three variables introduce complexities and, when managed poorly, instabilities. For example, too much process variation is an unstable foundation that makes it difficult to manage variety. Introduce unanticipated spikes in volume and the entire environment will be challenged to meet the demand. We often see numerous signs of failure under these conditions, expressed in terms of longer time delays, larger backlogs, increased number of defects, reactive crisis handling (“All hands on deck!”), decreased employee morale, and lower customer satisfaction.
What we’ve just described is the extreme case, but there are many service desk environments that suffer from mild forms of these symptoms over extended periods of time. Management and staff often just get used to this state of affairs, tolerating suboptimal performance in the name of getting by. There may be less urgency in this scenario, but the underlying workflow dynamics are the same.
Solving Problems
Now that we have a more nuanced understanding of these operational problems, what do we do next? The first thing is to understand each problem as it relates to a specific environment. Since we encourage our clients to look at each variable separately, let’s put together a narrative script that takes us from the statement of the problem and its symptoms to a description of some underlying causes of these problems. Then, let’s identify some approaches to problem solving, with reference to some methods, tools, and techniques that can be used to aid in remediation.
Our approach revolves around planning and managing these variables. We don’t want to be surprised by unforeseen changes in ticket volumes, for example, so that means we have to anticipate and manage volumes in a dynamic supply-and-demand formulation. There is such a thing as too much variety if a large number of tickets defy conventional (and planned) categorization, as we find in cases of exception processing (“I never saw that request before!”), so we need to explore ways to reduce unnecessary variety or define standardized processes that can handle a portion of the variety (i.e., a process for handling exceptions). The same notion applies to process variation. There is often much more of this than we see on the surface, so we need to make it transparent, and then we need to reduce variation through the application of standardized procedures.
The problem-solving script we use is a simple one that calls for understanding the problem using root cause analysis and then identifying appropriate solutions using methods, tools, and techniques derived from Lean management. Lean is a widely deployed and very flexible set of practices that are geared toward eliminating operational inefficiencies.
In the Lean approach, root cause analysis is more of a cultural orientation, one where we apply a form of scientific thinking to our operational issues. In other words, it’s not just a one-off exercise for a particular problem, but a way of thinking about problems generally. For example, we might be caught off-guard by spikes in volume, as might happen when there is miscommunication between parts of an organization (perhaps the marketing group introduced a special offer that resulted increased demand on our sales systems, but IT was not informed). When a pattern of spikes develops, we definitely want to understand why. The symptoms are going to be things like backlogs or bottlenecks, but there could be any number of underlying causes that we need to understand. Once we understand these causes, we can brainstorm solutions to address them. These solutions will likely include common-sense actions, like opening up a communication channel with marketing, increasing capacity, or increasing the flexibility of resources so they can be moved around to meet changes in demand.
As noted, several of these solutions are quite commonplace, and are likely to be solutions service desk practitioners are already using. However, these solutions may also leverage Lean tools like workload leveling, one-piece flow, 5S, and mistake-proofing, which are more likely to be unfamiliar to service desk practitioners.
What Makes a Service Desk Lean?
Lean operations can be accomplished through practice with the methods, tools, and techniques derived from the Lean Toolbox, but it also means adopting the philosophical orientation we associate with Lean Thinking. Lean Thinking involves the adoption of a problem-solving orientation combined with practices for continuous improvement—solving one problem, then another, then another, and so on, with a “practice makes perfect” attitude.
We define a Lean service desk operation as one in which volume, variety, and process variation are managed, and in which individual workflows are optimized for efficiency. That’s a very simple definition, and while it suggests a high bar for qualification, we believe that consistently applied and dedicated efforts to achieve these objectives are sufficient to warrant the label. In Lean practice, perfection is the goal, but getting there is where the learning takes place, and it is accumulated learning that distinguishes Lean operations from others.
* * * * *
Rebecca Duray is a professor in the College of Business at the University of Colorado at Colorado Springs. Rebecca received her PhD from The Ohio State University and her MBA from Case Western Reserve University. Prior to earning her PhD, Rebecca was a management consultant at PricewaterhouseCoopers and A.T. Kearney, among others. Her research interests focus on the strategic use of operations, process improvement, and Lean information systems.
Howard Williams is an IT service management consultant in Microsoft’s Services Consulting for IT Operations practice. He is an ITIL Expert and has an MBA in operations management from the University of Colorado at Colorado Springs.
Rebecca and Howard are coauthors of
Making IT Lean: Applying Lean Practices to the Work of IT (2012).