The service desk conundrum

BANGALORE, INDIA: The rapidly evolving technology climate has led to an end-user shift towards more sophisticated technologies. Employees are using more and more complex devices and applications at home and inevitably bringing some of these to work. They are also becoming more demanding than ever before and expect a higher level of support.

At the same time, CIOs are facing more financial oversight and are required to prove hard ROI on technology investments. This leads to more demand on their organizations to show demonstrable hard results and metrics. Traditionally, service desk performance has been measured and evaluated on operational criteria such as average handle time, first call resolution rate, and abandonment rate, among others.

In more recent times, the need for business metrics including cost, value, and end-user productivity have been widely recognized and adopted. Most service organizations today—whether they are employee facing, internal service desks, or externally facing customer service desks—are measured on increasing customer satisfaction as well as reducing overall support costs. And, more often than not, these two objectives are in direct conflict with each other. So what is a service desk manager to do?

A key metric used to measure performance across most organizations is first contact resolution. Since solving a customer’s problem is of utmost importance, a focus on this metric is not surprising. Yet what would result from an over-reliance on  first call resolution? If a service desk is measured solely on this metric and the analysts are driven to achieve higher rates, analysts would focus on:

• Keeping the end-user on the phone for too long, as the analyst worked on solving their problems
• Sub-optimal escalation rate, as analysts are trying to solve problems beyond their skill level in many cases—problems that should have been escalated to level two or a deskside visit for the quickest solution 
• Higher average speed to answer 
• Higher abandonment rate for calls

Such behavior takes the focus away from solving an employee’s problem, and instead focuses more on meeting the metric itself. The analysts are also less inclined to use the tools and technologies available to solve problems.

They are also less likely to encourage end-users to use self-service or help the organization deploy proactive services as these reduce the actual number of problems they solved directly—adversely affecting the metrics they are measured on. Let’s also delve into another commonly used metric—average handle time. A majority of service desk managers focus on reducing average handle time as the key to improving the quality of their service desk.

Similar to the challenge of over-reliance on first call resolution, an over-reliance on average handle time

can lead to:

• Unnecessary escalation
• Poor escalation—escalations without providing the right context
• Low problem resolution rates resulting in multiple
• calls to resolve the same issue
• Poor customer satisfaction

In isolation, both of these behaviors often lead to higher support costs and poor customer satisfaction, despite the best intentions on the part of help desk managers and executives to do the right thing. It’s simply a struggle between two seemingly competing objectives—reducing costs and increasing customer satisfaction.

The fact is, the traditional approach to metrics is a zero sum game. Also, many of these metrics evolved in the early 1990s when service desks were measured by operational metrics alone—focusing exclusively on level one productivity and their call handling skills. These metrics deliver an incomplete picture regarding service desk performance and don’t provide the appropriate guidance on best practices.

This begs the question—Is there a better way?

We are all well aware of the fact that incentives drive human behavior, so we need to evolve our current thinking and align behavior and incentives with accurate metrics to gauge real performance. What appeals to both end-users and analysts AND is very important to the service desk environment? It’s rather simple actually.

• End-users want to get back to work faster, and
• Service desk analysts want to solve problems more quickly.

Solving problems fast—this is one common and unifying goal across both groups. The other common goal is that both end-users and analysts want the process to be easy and seamless. Mean time to resolution can be one unambiguous criterion that could appeal to both groups.

Mean time to resolution is defined as the average time between incident reporting and its satisfactory resolution.

Let’s look at our first goal—resolving problems faster. If we measured how fast problems were resolved and measured service desk analysts on this metric, this would drive the analysts to solve problems faster. This would satisfy the end-user who would be able to get back to work faster.

Mean time to resolution would even appeal to the service desk managers and executives who care about keeping costs down while keeping customer satisfaction up. This definitely provides a more complete picture. Now let’s examine how mean time to resolution can be a far more superior approach to service desk management.

If the only metric of importance was reducing mean time to resolution, then analysts would follow the most efficient process for problem resolution. They would educate endusers to resolve problems themselves, use automation to solve problems faster, and use the most advanced tools and technologies available to accelerate the resolution process. With the obvious conflict—reduce cost versus increasing customer satisfaction—resolved managers and executives would be able to introduce the right technologies and processes while facing less resistance from analysts.

Analysts would then be able to clearly see technology and process improvement as catalysts to reduce mean time to resolution and would ultimately, perform better on the metrics they are measured.

In a recent survey of more than 200 service desk managers, we found that having instant access to accurate diagnostic information would make it faster (66 percent) and easier (62 percent) to diagnose and resolve problems. However, service desk managers aren’t necessarily taking the steps to ensure timely access to accurate diagnostic data.

• Nearly one quarter (24 percent) of IT help desk managers say that diagnostic information is keyed in manually into their call tracking system.
• Two-thirds (66 percent) of IT help desk managers say that it takes them six minutes or more to thoroughly gather diagnostic information about an end-user’s computer when an incident is opened, and one in three say it takes them eleven minutes or more.

While it is possible to get a high first call resolution in the scenario when you spend a lot of time on data collection, both end-user satisfaction and cost will take a hit. A large majority of service desks provide for multi-channel interaction—self-service, proactive service, and assisted  service. In most situations, all three forms of support are managed somewhat independently and end-users need to determine the right channel to get support. When an enduser moves across channels, the transitions are not seamless in most cases and the end-users have to provide context and information multiple times during an interaction.

To compound the problem even further, nearly all (98.5 percent) service desk managers say that if a call is escalated,  their level two and level three analysts have to re-input data the level one analyst already gathered. One of the key drivers to reducing mean time to resolution is identifying the most appropriate channel for the problem resolution and then seamlessly guiding the end-user to that channel.

Also important is the ability to let the user start in any channel and be able to seamlessly transition to another channel if they are not able to resolve their problem easily. Once the user is in the right channel, it is equally important to accelerate the problem resolution process using automatic diagnostic data capture, providing only contextually relevant solutions and enabling service desk analysts to solve problems faster, if assistance is required. 

An example of such an interaction would be an end-user trying out self-service to resolve a problem and trying out multiple fixes, but not being successful. If the user then  requested assistance via chat and none of the content from self-service was visible to the analysts handling the chat request, the end-user would need to provide the entire  context again. The risk is that the analyst would likely attempt all the same solutions the end-user may have already tried before reaching the right solution thereby, increasing  the time to resolution as well as frustrating the end-user.

Alternatively, if the self-service and assisted service applications were tightly integrated and the entire context including diagnostic data and attempted solutions were transferred to the analyst for the chat interaction, the resolution would be significantly faster and the end-user would be much more satisfied with both the outcome and the process. This would be similar to the experience with your phone company where you are pleasantly surprised (and almost shocked) when all the information you provided to the self-service voice response system gets transferred over to the customer service agent and you start wondering “…why can’t everyone else do it every time? This saves me so much time and reduces frustration.”

In addition, if the service desk solution helped deliver automated fixes for known common problems, the number of incidents would naturally go down and help reduce overall support costs. Having an integrated multi-channel service desk solution

would lead to:

• Reduction in mean time to resolution

• The end-user being much more open to trying selfservice in future instances, as they would have the confidence that the context would be transferred for assisted service

• Organizations successfully lowering costs as well as increasing customer satisfaction

The key to running a successful service desk operation is to adopt a combination of the right metric—mean time to resolution—along with the right technology integrated multi-channel service desk solution. By measuring and incentivizing your service desk analysts on the right metrics and providing the right technology to both analysts and end-users you would set them up for success. This would enable you to manage the seemingly conflicting goals of increasing customer satisfaction and reducing support costs.

About the Author 



Rajeev Shrivastava leads product marketing and strategy for SupportSoft’s enterprise business globally. His responsibilities include new product introductions, positioning, pricing, partnerships and market expansion.