Datapro Summary

The help desk can be both a boon and a burden to a company, either increasing customer satisfaction and return on investment or making customers and technicians miserable. One of the keys to achieving the former is selecting a help desk package that not only fits the company's present and future needs but is designed to make the technicians' jobs more manageable and less complicated. There are over 200 software products available, ranging from simple call tracking packages to more sophisticated problem resolution and expert systems. To prevent any of these from being a burden to the company, the IT manager must understand and consider the company's business, its networks, and users to select the most appropriate help desk package. By John F. Inverso

Technology Basics

The help desk's primary mission is to provide a single contact point of end-user support within an enterprise. Functionally, the help desk also provides general systems information, responses to requests on procedural matters, and logging of all problems; it has--at a minimum--first-level responsibility for problem determination and system restoration. Help desk software should be able to provide the following features:
· Logging problem calls.
· Providing one centralized contact point.
· Collecting product evaluations from end users.
· Helping to resolve problems quickly and efficiently.
· Tracking and managing difficult problems.
· Improving the productivity and training of end users.
· Providing a way to escalate very serious problems and those that are slow in being resolved.
· Identifying recurring problems, reporting trends to management, and recommending solutions or procedures to correct faults.
· Resolving as many problems as possible on the initial call, thereby reducing the number of calls redirected to upper-level personnel.
· Supporting and managing configuration changes plus inventory tracking in support of information services.
· Generating and using problem history to improve the availability of various systems and equipment.
· Developing standard management reports for evaluating vendor performance and service-level contracts.

Help desk systems generally fall into two broad categories: instructional tools advising help desk staff about what to do and information retrieval tools providing useful data but leaving interpretation to the user. Each method features attractive attributes, so the best approach depends upon specific organizational needs.

Instructional tools basically tell the help desk staff what to do. These tools lead the operator.to solutions using logical analysis chains, culminating in problem resolution. The critical component of information retrieval tools is the knowledge base, which contains specific data about particular subjects. It can contain troubleshooting strategies, device descriptions, and the expertise of many technical resources. Help desk information is accessed using simple scripting approaches with decision tree logic or with complex expert systems using rule-based reasoning.

Retrieval tools, on the other hand, present information that allows individual interpretation of data but require more skill to use. There are three major types of retrieval tools: text search, hypertext, and case-based reasoning. Hypertext lets the user search large volumes of information by linking phrases and keywords similar to word and text searches, but much faster. It eliminates the text search problem of "too much data" when using nonrestrictive search arguments. Case-based reasoning is a broader retrieval approach, replacing text and keywords with case-by-case situation searches similar to human experience. It requires complex indexing algorithms and sophisticated software; entering additional information is just a matter of data entry. Its primary advantage is the capability to learn from experience, making every new case available for future access.

The best help desk solution may be hybrid systems combining several methods in different situations, such as linking keywords with cases and presenting multimedia information. Icons and pictures can represent equipment and configuration schematics, easing problem analysis and determination. Implementers must remember, however, that every help desk system and tool must log, track, and retain information in a data repository. In this way, help desks can become proactive, anticipating and predicting certain problems while collecting and reacting to new information and situations.

Help desk tools and utilities vary widely in features, hardware/software platforms, target markets, cost, and functions. The basic elements of help desk automation tools include the following:
· Call tracking and management--The foundation of any automation system, identifying and classifying every help desk service request are the most basic help desk tasks. Call tracking data helps managers understand what their systems and networks are actually doing, as it enables them to spot performance trends and isolate problem areas. Meanwhile, call logging, call detailing, and call assignment create the decision-making infrastructure needed to make rational configuration and implementation choices. Call tracking systems must include report generation facilities that classify every received call.
· Problem solving--Necessary to improve both technical accuracy and speed. This feature can vary from product to product from simple text searches to expert systems using artificial intelligence.
· Escalation--Often automated, this feature enables a call or task to be forwarded to a technician for further action. This usually occurs when a call cannot be resolved by the help desk operator or when a specific action needs to be taken that requires a technician (i.e., changing toner cartridge in a printer). Escalation rules are stored in the products database and can be changed by authorized personnel.
· Call retrieval--Provides a history of all calls logged into the help desk. This feature permits searches for particular information or statistical features that generate specialized reports.
· Information retrieval--The next logical step following the administrative functions of call and symptom logging. Information retrieval always occurs during the first attempt to resolve a call, whether or not this function receives assistance from artificial intelligence. The most common approach is using the database's normal search engine to find historical records, demonstrating how similar problems have been resolved. Though useful, this method has one hitch: simple database searches usually retrieve numerous hits and do not designate the most pertinent ones, so the analyst must browse through them individually.

Statistical searches can prove more useful at sifting through the large volumes of text likely to be found in a call database. A statistical search counts the number of times a search phrase is.encountered and then ranks the records by their relevance to the search criteria. In addition, these search engines let an analyst exclude "noise" words from the search-and-specify synonyms:
· Reporting--Customized report writing capabilities, using the database to extract and identify type of calls received, system trends, and other low-occurrence rate information.
· Databases--Configuration and equipment inventory databases, listing hardware and software installed in the network. These records must contain appropriate vendor information--names, addresses, emergency data, etc.--and problems associated with that equipment.

Generally, systems featuring windows-based graphical user interfaces frequently let users hot key into graphics libraries containing images of proper hardware switch settings, system configuration diagrams, specification charts, and floor maps. This capability helps analysts more quickly diagnose and describe problems, enabling technicians to more easily locate and resolve problems. Other features of advanced help desk systems permit operators to perform the following:
· Use multiple variables to search call records for similar problems.
· Establish a single file to record multiple calls caused by a major system failure.
· Audit changes made to call records.
· Automatically dial phone numbers (home and business phones, beepers, cellular phones) to alert maintenance and support personnel.
· Automatically e-mail appropriate personnel of the problem, its status, and impact. Increasing emphasis is being put on the development of features that extend beyond the basics of the typical help desk product into the realm of enterprise management. This development of consolidated service desk software has enhanced the existing features of the help desk and added a host of abilities that include the integration of various network and asset management features.

Expert Systems
Artificial Intelligence (AI) or expert systems basically interpret partial or incomplete service request information and recommend logical solutions. These systems generally consist of rules-based or case-based reasoning. Expert systems infer the most probable cause of a problem or most probable answer and recommend solutions. Responses are derived from previous technical experience, which is made available following rules generated by knowledge engineers--individuals with extremely broad and detailed technical knowledge. Case-based systems, sometimes called neural networks, determine solutions by comparing previously entered examples and problems. The case-based system changes with each problem entered and solved, providing an intuitiveness sense that the system learns from experience. Developing either type of expert system requires extensive knowledge, engineering, and programming skills, and can require an extensive implementation period for complicated environments.

When installing an expert system, experts from all technical areas usually work at the help desk to construct a software answer base, thereby helping train less technical support staff to diagnose and resolve future problems. The experts or developers use the system's fields for actions, objects, and topics to select and describe potential problems and their solutions. Transaction files store these answers. At the end of each week, a high-level analyst reviews the files and deletes inappropriate or incorrect answers and then routes the reusable information to the main database. Novice help desk personnel use this answer base to guide callers through simple diagnostic sequences, beginning with the problem's description and then progressively trimming down the results until they reach a plausible solution.

Automating the help desk function improves service quality while decreasing its costs. It also frees up experts to share their experience with other departments, promoting a consistent answer set throughout the company and improving the help desk's credibility. Some expert systems use the database to generate graphics and textual reports on the hardware and software that cause the most problems. Future purchase and design support systems then use this collected information.

Automation
One of the most widely developed aspects of today's help desk products has been their ability to free the help desk operator from as much time-intensive work as possible. By automating various processes, the user is able to address more calls rather than spending valuable time logging information and manually searching for the proper resolution. Automated tools and utilities, however, are useful only if implemented and used properly. Selecting the appropriate software system, installing it, and managing its functions and training support will yield maximum benefits but require careful planning and implementation. The help desk should also integrate network problem management strategies and anticipate network growth. Today, systems and network support centers continue to grow in importance as information system and business priorities evolve. Automation tools available in this area include the following:
· Displays that contain problem determination information. These include AI-based applications offering device/procedure-specific multiple-choice questions that let the help desk staff and/or end user determine the problem's cause and resolution. Web browsers and other groupware enable end users to access the knowledge tree and other artificial intelligence utilities so they can resolve the problem without submitting a complaint or trouble ticket to the help desk.
· Dynamic, realtime summary displays. These are terminals or overhead displays that project meaningful information on the current system status. They include device monitoring facilities that continually track network performance levels and proactive equipment to forewarn help desk personnel about impending network crashes, etc.
· Aids to automate routine help desk tasks. These include tools that log routine/identification information about the device/node that is calling the help desk. Help desk staff can then gather data about terminal/set ID, caller's ID, location of the entity in the network (LAN ID, etc.), time/day/date of the call, most recent call from the same source, etc., using computerized support.
· Multisession window aids. Such facilities let help desk personnel move from session to session without logging off one session and logging on to another. These navigation tools enable the help desk staff to "make a backdoor entry" into the user's onscreen procedure and mimic the session to pinpoint the source of error.
· Statistical aids. These facilities generate a variety of customized reports on the data gathered at the help desk.

The actual degree of automation depends on help desk staff itself, including its familiarity with the current system, its capability to manage automated facilities, and the projected improvements in service levels and call volumes handled. While there are no clear-cut answers, the preceding guidelines are customizable to the implementation needs of the organization.

Web Interface
One of the most recent advances in help desk software is the expansion of its functionality into intranets and the Internet. The extent of this functionality varies greatly from product to product. This provides help desk personnel, the help desk clients, or both with access to various aspects of the help desk system. The help desk operator or technician, using any standard Web browser, can access, edit, and update call records, call status, a host of reports, and any other feature offered in the help desk package. Some products even go as far as reproducing the user's help desk interface on the Web browser, making it look exactly as it does on his or her desktop.

For clients, some help desk vendors designed their Web interfaces to provide a "help yourself desk" feature to reduce the amount of calls to the help desk. Clients should be able to access the knowledge base and its search capabilities, allowing them to query the database and view resolutions to some of the more common problems without requiring operator intervention. Clients should also be able to log calls from the Web browser when actions are required that require a technician. These call records are automatically processed and escalated to the proper.technicians via the software's rules set and automation features.

Desktop Management and Administration
Desktop and network management functions represent another trend in the help desk market. Customers are demanding ways to more efficiently manage their networks, especially their desktops. Each desktop costs about US$10,000 a year to support. The help desk is typically the first place an end user calls when experiencing problems, and it often authorizes the necessary changes and maintains responsibility for the resolution. Asset management features enable the help desk to view the current information of any caller's hardware and software assets. This is key to diagnosing and resolving problems related to the asset itself. This can also be used proactively to determine which assets need upgrading or replacing before conflicts arise. Adding change management features also help the help desk analyst manage the process of changing current network and desktop configurations and maintain accurate documentation. Ideally, the relational database driving the help desk application should automatically update all corresponding records when a change is made to resolve a problem or enhance the network. Many leading vendors currently offer asset, change, and workforce management applications as add-on modules to their help desk systems or have announced plans to do so.

Management System Integration
Rather than relying purely on a reactive process of problem resolution, many larger companies are attracted to help desk products that can utilize the information from their network and systems management platforms. This proactive model uses the diagnostic and reporting capabilities of the management platforms to alert technicians of potential problems. Typically, the management software would detect a potential or real problem (i.e., service-level violations, asset or application performance degradation, etc.) and report it to the help desk in the form of a trouble ticket. It could then be handled like any other call. Those products with sufficient call escalation features (usually customizable) could send the ticket to the appropriate technician, if so specified in the help desk's escalation rules. Also, all call and problem information could be recorded for further process and system analysis.
This level of integration serves to consolidate the management of the company's customers (internal or external) and of its network and systems services into a centralized point of control. Administered correctly, this could add significant value to the help desk system through faster first-call resolutions, proactive problem identification and resolution, and increased awareness of recurring trouble spots which could then be fully addressed.

Help Desk Tools
As the help desk moves into a more prominent and valued position within many companies, software developers have invented tools to make problem resolution and administration easier. Close to 200 help desk tools and packages currently compete in this market segment, offering varying capabilities, features, and pricing from simple call tracking systems for a few hundred dollars (U.S. currency) to sophisticated expert systems costing tens of thousands, depending on the size of the configuration and user base. These tools include:
· Remote Control--let help desk analysts remotely access the end user's desktop so both the analyst and the end user are viewing the same screen at the same time. The support person can even seize control of the affected desktop when necessary and guide the user through the problem-solving process by entering appropriate commands. This method provides an excellent vehicle for quick and efficient troubleshooting and makes it possible for organizations to centralize their end-user support functions. Therefore, technical support people no longer have to.run from location to location to fix problems. Instead, they can see exactly what is happening at the caller's desktop in their own offices. Some of these tools run over LAN connections rather than modem connections, enabling authorized stations on a LAN to view each other's screens and control each other's activities.
· Session Recording--special utility programs allow users to record interactive terminal sessions or portions of sessions in log files. Should users experience problems with applications, they can send their log files to the help desk staff, enabling technical support personnel to examine problems without having to first reproduce them. This is especially useful when racking the source of an intermittent problem.
· Remote Device Access--using remote communications tools, the help desk operator can freely access the different devices on one or more LANs. A LAN connection is much faster than a 2400 bps modem connection and lets users jump around, view, and control a variety of equipment. The help desk operator can carry out all the maintenance activities that once required an onsite technician.
· Status Alarms--more advanced products let help desk operators set alarms as a reminder to check a problem's status. The help desk can suggest alarm times based on the call's type and priority, and can escalate alarms to a technician's supervisor or manager when appropriate. Many of these more sophisticated systems feature knowledge trees, case-based reasoning, artificial intelligence, and problem resolution capabilities.

Operating Requirements
Operating requirements vary greatly from product to product. There is a help desk product available for most any mainframe and client/server environment currently in use, including multiplatform and distributed environments. Many of these products integrate at different levels with most network management platforms and frameworks, such as HP OpenView, Intel LANDesk, CA-Unicenter TNG, Tivoli Enterprise Software, and Cabletron Spectrum, etc. But there are also standalone packages available that can require as little as a single client and a server. To specify a set of requirements would be impossible, since there are enough options on the market to address almost any network configuration.

Business Use

Most any company that has a computer network can benefit from some form of help desk software. At the very least, it provides the company with a means of organizing and managing calls to the IT department for help. At its best, the help desk can be a cost-saving implementation that not only handles client requests and queries but also acts as a storehouse of information about the network's overall performance, reducing the total cost of ownership of network devices, applications, and services. It can also function as a valuable on-the-job learning facility for network users and management personnel.

With the growing complexity of problems associated with client/server-based networks, companies are not only finding a greater need for problem and knowledge management, but for the management of the internal and external resources that affect their overall operations. Many of the high-end help desk products offer methods of tracking and managing the service-level agreements of their own IT departments as well as any contracted support services that must.maintain certain levels of performance and expediency. Likewise, larger companies can benefit from the "help yourself desk" capabilities of some Web-enabled products. Often, too much time is spent by help desk personnel dealing with routine calls and by end users waiting for responses from analysts or technicians. The "help yourself" concept inherent in Web interfaces allows the analysts and the end users to access various parts of the product, depending upon their authorization. Help desk personnel would be able to resolve problems from any Web browser while end users could log calls or access the knowledge base on their own, thus enabling them to get back to producing more quickly. These features, along with the host of other capabilities of some of the more advanced help desk products, are worth the investment when the time-, production-, and cost-saving benefits are carefully considered.

Benefits and Risks

Substantial Savings Over Time : Putting a dollar figure on help desk benefits can prove difficult. The network's importance to each organization varies within a business, and though vital, its benefits are often intangible, at least in the short term. However, when compared to the costs of network downtime or simple poor performance, these benefits should shine brightly. Network failure or reduced availability, even for a few minutes, can cause disastrous losses in major financial organizations. Here the money spent on help desk operations that speed problem resolution translates into substantial savings.

Avoid Impulse : Shopping While help desk software can benefit a company's bottom line as well as its network performance, not every package will yield the same results. A single-site organization with under 100 users and one or two help desk operators would be foolish to spend tens of thousands of dollars on a fully automated help desk solution. While many of today's packages have very attractive features, an organization needs to carefully distinguish between features they would definitely benefit from and those they can do without. Likewise, a larger organization with multiple sites, thousands of users, and an entire help desk department would hardly find much benefit in a simple call tracking package. An organization must consider the number of analysts it is trying to support as well as what its projected growth will be. Only after detailing its current and future needs can the company better invest in a product that will not largely be left unused.

Standards

There are no set standards for governing help desk software, although proposed standards were released in 1998 by DMTF. These proposed standards include SIS (Incident Exchange), which deals with exchanging trouble ticket information, and SES (Resolution Exchange), which is directed at the exchange of resolution knowledge. The industry is reluctant to adopt these or any standards until there is a significant user demand for them.

Price vs. Performance

Entry-level systems are usually simple call tracking systems that will support a limited number of users and transactions. Some vendors have gone as far as offering such software for free download from their Web site. Other vendors offer limited call management features. These types of products are usually "out-of-the-box" packages that are intended for fewer than 10 users. As such, these products can range anywhere from a few hundred dollars (U.S.) for the most basic.package to around US$15,000. However, should the customer grow, many vendors offer upgrade modules and licensing options to scale their products effectively. Making up the mid-level are the products with these added features included. Typically, they will feature some automation and advanced search options as well as customization. The high-end help desk products are for the large-scale environments that need to track not just calls but performance issues, service levels, and a host of other factors. Beginning around US$50,000, these systems often integrate with network management platforms; support a large number of concurrent users; and offer detailed information-retrieval systems, such as expert systems and Web interfaces. The cost for this added functionality is worthwhile if the customer desires high levels of call resolution and performance reporting and management capabilities.

Selection Guidelines

When selecting help desk solutions and devising a strategy, management should examine long-term organizational goals, current and future business functions, and anticipated network growth and usage. Once implemented, the help desk system should grow with the company and its network, alleviate most user frustrations, and demystify network complexities. Achieving this ideal, however, is not always simple and requires well-trained help desk staff and informed network managers as well as a good software system.
Some experts believe it is very difficult to solve all user problems at user expectation levels. Users usually become dissatisfied with their help desk when they misunderstand its functions or have unrealistic expectations. Help desk staff cannot answer every question every time. It must have a cavalry behind it armed with escalation capabilities and access to complete resources, both internally and externally. To meet user expectations, management must clearly define help desk roles before installing the network. Managers must also determine beforehand whether the help desk should mainly "fight fires" by resolving immediate user problems and/or also function as an important data gathering center for enterprise management. Managers need to define the following criteria:
· Types of problems the help desk must solve.
· Reporting procedures for referring long-term, unsolved, difficult problems.
· Types of data gathered about network performance.
· Report generation formats.
· Help desk staff responsibilities.

Help Desk Staff
Properly selecting and adequately training help desk staff cannot be overstressed. "Help desk staff need not be technical experts and must communicate in a jargon-free language," says Ronald Mums, president of the Help Desk Institute. He also believes that most help desk personnel, when properly guided and managed, can soon master the facility's technical aspects..
Unfortunately, help desk personnel often suffer from stress and the unpredictable work pace. They feel unappreciated by both the people they help and their own management. On occasion, they do not understand what is expected of them. It is not unusual for a help desk person to "know what is wrong without knowing how to help the user solve the problem." Help desk personnel face multiple challenges that include identifying the user and the user's location, locating the faulty network device, determining whether the problem is chronic, and determining the status of the procedure in question.

Automation Needs Careful Planning
Automating the help desk solves many problems, but it is not a panacea. Automatic solutions take time and effort to build, test, and implement. In fact, unplanned automation will only increase problems and create a gap between the users and the help desk. Automation should contain and address the following general characteristics:
· Overall network management goals.
· Some form of problem resolution.
· Inventory and asset management.
· Change management.
· Support for core business objectives.
· Capability to customize without breaching the integrity of the existing application or the data it maintains.
· Security capabilities to provide read-only access to configuration databases.

Technology Leaders

Goldmine (formerly Bendata)
One of the first PC/LAN-based automation tools designed for help desk management, HelpDesk Expert Automation Tool (HEAT) was designed by the joint efforts of The Help Desk Institute and Bendata Management Systems, Inc. HEAT enables help desk operators to access call logging and call description, assign problems, journal the entry of activities, and close problems with a single keystroke operation from a console PC. HEAT for Windows automatically routes calls to the first available help desk analyst and uses a knowledge tree to search for answers and sorts them according to frequency of use. The Manager's Console monitors the system and recommends corrective actions when thresholds are reached. It also capitalizes on the 24x7 management philosophy and e-mails open tickets to an available help desk in another time zone when the participating one closes for the day. The system features a service-level agreement function that monitors service levels. HEATlink to the Internet lets users submit problems and check trouble ticket status. HEAT also provides for a customized solution database, immediate access to caller profiles, configuration and call history, and other features. The system can also integrate with enterprise management systems, such as Hewlett-Packard OpenView, and Tivoli TME.

Clarify ClearSupport and ClearHelpDesk
Clarify supports external and internal help desks with its ClearSupport and ClearHelpDesk products. ClearSupport provides problem management software for external support organizations. It automates call tracking, escalation, work orders, and expenses for support staff.based onsite and in the field. Designed for the internal help desk, ClearHelpDesk features similar call tracking capabilities as ClearSupport and also manages service-level agreements. ClearHelpDesk can also interface with leading network management systems, including HP OpenView, SunNet Manager, and Microsoft Systems Management Server, and can integrate with off-the-shelf knowledge bases from KnowledgeBroker and ServiceWare. Clarify's Diagnosis Engine acts as a proactive knowledge base by collecting and storing problem information and using case-based reasoning to help staff troubleshoot problems.

Computer Associates Intl. ServiceIT Enterprise Edition
Computer Associates markets two help desk products, SerivceIT Enterprise Edition and Unicenter Advanced Help Desk (AHD), which utilize a similar core technology but target different market segments. ServiceIT is a standalone product targeted at customers who need to address external service center management, while AHD is integrated with Unicenter TNG and addresses internal or enterprise service center management. ServiceIT combines call management and problem tracking capabilities to provide service desk staff with an integrated solution. Both systems capture problems that are either manually input or automatically generated from network management platforms or agents. They combine an asset inventory application with a capability for modeling the associated systems and network management workflows and processes. They also generate trouble tickets, notify service staff and end users of a problem's status, and escalate problems until they are resolved. Paradigm supports Unix and Windows NT operating environments and Windows 95, 98, and NT operating systems, as well as TCP/IP protocol and several relational databases.

Network Associates Magic Total Service Desk Suite
Originally from Magic Solutions, Network Associates acquired this suite of help desk products when it acquired the vendor. This suite offers the typical help desk functions along with several features to extend it into the area of enterprise management. The Total Service Desk Enterprise Edition includes asset, event, and desktop management and remote control features. Its sister product, Magic Help Desk Enterprise Edition is entirely browser-based and provides call, problem, and knowledge management features with an emphasis on reducing user-support expenses. Its browser-based architecture lends to the product's reputation for being highly scalable and customizable. Other modules in the suite can be added to provide greater customization options, database management, and software metering and distribution. The reputation of these products have earned the vendor some high praise, especially for its browser-based product that has promise and reflects the possible future face of help desk software.

Remedy Corp. Action Request (AR) System
Remedy Corp. originally designed the AR System as a network-based messaging and active notification tool for automating internal help desks and recently expanded its focus to encompass "consolidated operations management" (COM). COM includes automating other business processes, such as asset, change, and defect management. The AR System features an "experience database" that forms a complete diary of problem resolutions. Therefore, it can satisfy more than one need by structuring intraorganizational messaging and workflow processes while its open architecture makes it easy for customers to modify the system. As a result, Remedy enjoys a successful repeat business as current customers purchase more systems and the new companion products to automate other business operations, prompting it to develop its COM approach.

Technology Alternatives

The call tracking capabilities of the low-end help desk products can sometimes be handled through similar capabilities in Microsoft BackOffice and Lotus Notes. However, these will not support the needs of customers desiring the added functionality of the midrange and high-end systems.

Datapro Insight

The help desk market is evolving and using a more proactive than reactive approach. Many middle and high-end vendors are enhancing their trend analysis and problem detection and resolution capabilities so help desk staff can foresee a network problem before an end user experiences it. Vendors are also upgrading user interfaces and transmission media to make it easier for users to report problems and for help desk staff to retrieve and respond to them. The demand for these capabilities and those beyond that of the traditional help desk will continue to grow as new technologies bring higher levels of access and automation to the network infrastructure. The result of this growth is already becoming apparent as more of the marketplace is realizing the connection between competent help desk products and a more appealing total cost of ownership regarding the networked environment.