The service desk as many of us know it is at a crossroads. Specifically, service desks need to dramatically change their current approach to incident and problem management.
Many IT organizations are fighting a losing battle against growing IT complexity. As cloud services, devices, and data proliferate, it’s often difficult for IT to even get a grip on what it’s managing, much less resolve problems efficiently. These trends inevitably converge and a big mess lands on the service desk when something isn’t working.
Unfortunately, service desks face a fundamental challenge as well, in that most are set up according to a very structured process: an incident comes into a centralized area, an agent or technician categorizes the issue according to existing definitions, and then the incident is assigned to the appropriate group within IT. As service delivery becomes increasingly decentralized, this structure does not align well.
Service desks need a more advanced approach to incident and problem management. Big Data techniques are already beginning to transform overall service management, allowing organizations to exploit unused data trapped in their service desk databases and mine social interactions for greater intelligence.
As cloud services, devices, and data proliferate, it’s often difficult for IT to even get a grip on what it’s managing, much less resolve problems efficiently.
Ingenuity is turning the challenges of IT complexity into opportunities to dramatically improve performance. With Big Data, support organizations can deliver improved service quality and radically reduce the number of incoming IT tickets. Big Data intelligently connects the IT landscape with user interactions in a way that leverages collective insight, improves decision-making, and helps your IT organization identify and solve problems faster.
Exploiting Unused Data
Data is both a problem and a solution. There was a time the challenge was in collecting data; now it’s in working with it effectively and efficiently.
This is playing out in various ways. In medicine, for example, researchers are recognizing the immense wealth of untapped information in hospital laboratories, as this New York Times article explains:
Large, costly, and time-consuming clinical trials are rarely carried out for uncommon complications…In the absence of such focused research, doctors and scientists are increasingly dipping into enormous troves of data that already exist—namely the aggregated medical records of thousands or even millions of patients to uncover patterns that might help steer care.
It’s an interesting analogy to the service desk, which has historically relied on structured data. Despite the various evolutions of user interfaces, programming approaches, or even delivery mechanisms, service desk reports continue to be defined, run, distributed, and reviewed in weekly staff meetings, year after year, relying on the same founding paradigm that assumes we know what we're looking for. The underlying data is structured to answer traditional questions. However, this approach has lead to only modest gains in terms of productivity, even as technology has progressed.
A more productive approach is to exploit the vast quantity of unused data trapped in service desk databases. This data often spans:
- Incident descriptions
- Questions or agent discussions
- Other text fields
- Attachments and various user interactions
The benefit of this approach leads to tangible benefits for all of a service desk's various stakeholders: the business user, the agent/analyst, and the manager/supervisor.
A key pain point for business users interacting with the service desk is that they have to provide a lot of information—sometimes more than they're even capable of providing—because the service desk relies on structured data to triage, categorize, and route incidents. There are two consequences:
- Business users are annoyed by having to enter information because it's time-consuming and, most likely, incomplete or incorrect
- Incorrect information leads to miscategorization that requires rework by agents and technicians—and, in many cases, creates erroneous reports because the data is wrong
By properly applying Big Data analytics, service desks can free business users from entering this unnecessary information, and instead request that they only supply a minimal amount of information. For example, the University of Pretoria, the largest residential university in South Africa, reduced the number of clicks required to submit a service desk request from eleven to one: simply attaching an error message and clicking “Submit.” The service desk can now learn intelligently—and correctly—how to deduce the rest of the data that is required for classification and further processing.
Big Data techniques also enable self-service, in a couple of ways:
Big Data turns social IT into business knowledge: When combined with social IT management, Big Data allows organizations to capture user interactions and create knowledge from conversations, chat, email, and other media. This grows your organizational knowledge from a unique business user perspective, and it enables users to access and share IT knowledge so they can help each other and themselves solve their own problems, rather than submitting a request that sits in queue.
Big Data enables deeply embedded knowledge: With the right analytics tools in place, Big Data allows you to combine structured and unstructured data from multiple sources with relationships and thereby expose embedded knowledge with additional business context. This is key to informing an effective search function that delivers meaningful and relevant results—the gateway for users to extract knowledge from Big Data.
Through the effective application of Big Data, self-service becomes a reality as it creates more opportunities for users to address issues directly and reduce the number of tickets in the service desk’s queue.
Agents and Analysts
Of course, self-service isn't the solution for all incident and problem management. Despite various improvements in different aspects of the service desk, the major obstacle to achieving productivity gains remains the same: service desk agents and analysts don’t know what they don’t know. Let’s look at two specific examples.
Access to contextually relevant data
When dealing with a particular process artefact (e.g., an incident), the efficient handling of that process artefact often requires access to another process artefact (e.g., a change) that relates to the same information, but not in a structured way. A structured approach would require a carefully planned definition of the data model so that the connection between the two artefacts can be designed upfront. More commonly, the association is manually made by the agent across the process artefacts.
Practically speaking, the agent or analyst needs to be able to see what information in other parts of the service desk is related to the part he/she is working on in order to be able to make a connection and boost his/her productivity. In other words, a smarter system provides visibility and insight to process incidents faster.
Through the effective application of Big Data, self-service and social collaboration become a reality.
Initiating problem isolation
The biggest issue in the problem management process is identifying the problem’s root cause and deciding where to start. This typically requires a combination of guesswork and tribal knowledge, which is an inefficient way to work.
Inside any service desk lies untapped information, hidden in a sea of incidents, that's sufficient enough to provide hints on where to start the problem isolation process. Agents and analysts can replace their guesswork and hunches with a data-oriented approach.
Consolidating and leveraging fragmented pieces of knowledge sprinkled across the business allows IT to create a more efficient and productive service desk. For the University of Pretoria, issues are immediately apparent, which means the service desk can assist clients quicker. By shaving 30 seconds off of every service desk interaction, the University of Pretoria saves 670 working hours per year. By looking at incidents and reducing them by a minimum of 10 percent per month, it can save 1,536 hours, which equates to a free day per month for each IT team member.
Helse Midt-Norge IT (HEMIT), the IT service provider for nine hospitals in central Norway, experienced similar benefits by implementing Big Data analytics to its service desk. In recent years, as the number of incidents climbed, it became clear that problem management’s reactive approach was too slow. With Big Data analytics, HEMIT was able to spend more time solving user problems, instead of just writing about them. The result was a 50-percent reduction in time to complete service desk self-service tickets, with 10 percent fewer phone calls to the service desk and 25 percent more service desk resources available for incident resolution.
At HEMIT, improving ticket quality improved incident management, resulting in a shorter time to resolve. By removing what doesn’t matter, Big Data analytics simplify the interface for both the service desk and the user.
Big Data is also able to capture ongoing user questions and searches, showing what’s occurring right now, so service desk agents and analysts can quickly identify patterns and submit a problem. People in the IT organization can then isolate and fix the issue, so that others don’t encounter it. This proactive approach improves service quality and customer satisfaction, while staving off calls to the service desk.
By tracking the topics users are searching for, IT can identify business trends and react. If, for example, many users are searching for articles on a new device that isn't supported, IT may want to add articles that end users can access, offer the new device as part of their service catalog, or simply add established support guidelines for BYOD. All of this improves customer satisfaction and the productivity of both users and service desk staff.
To this point, we've focused on how Big Data can improve the relationship between the business user and service desk agents and analysts. But let’s step back and assess the value Big Data can have at the level of the overall process.
In recent years, IT has been unable to substantially and measurably improve the average maturity of IT processes. Analyst firms such as Gartner and Forrester have defined several process maturity models (usually divided into five stages of maturity), but in the last decade, the overall industry has stagnated at a disappointing average maturity level of 2.4 out of 5.
Many service desks do a rather decent job of reporting on process KPIs as defined by frameworks like ITIL or COBIT. But the train stopped at the KPI station and hasn't moved further.
One of the reasons for this disappointing result is that the increase of maturity has traditionally been predicated on initiatives that are extrinsic to the service desk system, including process reengineering projects or other similar forms. Such projects have not borne fruit because, among other reasons, the time scale of such projects is often out of sync with the technology refresh cycle.
Many service desks do a rather decent job of reporting on process KPIs as defined by frameworks like ITIL or COBIT. But the train stopped at the KPI station and hasn't moved further. However, a typical service desk does, in fact, have enough data to suggest how those processes can be improved. The challenge is unlocking the untapped actionable information in its systems and exploiting it.
A New Approach to Service Management
The key is to put Big Data techniques to work and transform an area that has not evolved away from its founding principles. It needs to reinvent how users, agents and analysts, and process owners (supervisors and managers) relate to the IT workhorse and foster new levels of engagement and productivity.
A new era of IT services is emerging, and your support team and end users will require instant access to huge amounts of structured and unstructured data to gain the actionable insights they need to resolve issues. When empowered with Big Data, social collaboration and self-service helps to deflect tickets and alleviate the burden on often-overworked staff, freeing up IT resources so they can focus on more value-added activities. Advanced functionality will help agents quickly prioritize and categorize issues—and then assign them to the right domain so that the problem is resolved the first time out.
Jacques Conand is product director for ITSM software products at HP, where he has overseen the development of ITSM and ITAM software products for the last seven years. Jacques has more than twenty years of experience in the software industry, from equipment manufacturers to software vendors, in various roles, including software engineering, product management, and business management.