Introduction.  Help services do not just appear from nowhere and immediately provide useful services.  It is essential to be able to organize and manage them to provide worthwhile support to end users.  It is almost always the case that this requires a more formalized structure such as a helpdesk.  A helpdesk is a place, physical or online, where end users can get answers to their computing questions, report problems and request new services.  The helpdesk staff needs to be able to deal with the daily problems that are part of every computing infrastructure.

It is also usually worthwhile to have sys admins associated with the helpdesks.  For example, should improve both the proficiency of the helpdesk's staff and ensure the sys admins have first hand familiarity with many of the recurring difficulties the end users are encountering.

Self-help systems are also important.  Though they should not be used to entirely replace systems based more on human interaction for any number of reasons.  Fortunately, they can provide some worthwhile supplements and ease congestion.  It is often very worthwhile to provide at least some repositories of documentation either online or in places like computer labs.  Lists of FAQs - Frequently Asked Questions are quite common on websites.  These lists save support staff considerable time by answering relatively routine questions.  They also provide end users an opportunity to avoid hassles with central computing staffs.

The helpdesk should have a customer service orientation.  Ideally, it should be readily accessible and pleasant.  The tone set by the staff will do much to determine the the level of interactivity.  A willingness to help end users can go a long way toward mitigating other difficulties.

In these veins, it is also going to be essential to ensure there are adequate staff.  Determining the level of staffing by the ratio of end users to staff can be helpful, but also misleading.  It is more important to focus on call volume ratios and times to completion.  Unfortunately, these can also be misleading.  For example, a high call ratio with fast completion rates can really be the result of a failure to fix something at a slightly higher level of the system.  Maybe initial e-mail profiles weren't configured correctly so that every individual user needs to visit the helpdesk with an easily fixable problem that really should be fixed in other ways.

But, call volume ratios and completion times are more likely to represent customer demands than many other assessment measures.  It is also important to have credible performance metrics in order to impact higher level management's decision making.  The more objective the data seems, the more likely they are to consider it.

Defined Scope of Coverage.  One of the most irritating and difficult things to deal with for end users is the scope of coverage of the helpdesk, assuming the helpdesk is at least capable of some worthwhile coverage.  The end user can fell like a ping pong ball getting bounced from authority to authority while wondering what else these authorities are actually capable of doing.  But, irregardless of this the managers of the help desk are likely to need to consider the following issues.

  • What
    • What is being supported?
      • PCs
      • Network
      • Laptops
      • Telecommuter's computers
      • Particular operating systems
      • Particular applications
    • How are unsupported platforms dealt with?
  • Who
    • Who will be supported?
      • Particular departments
      • Particular buildings
      • Particular sites
      • Users that work at multiple  locations
      • People that pay
      • People with particular organizational clout
  • Where
    • Where are the customers?
      • On site
      • Traveling
      • Consortium related users
      • Sales personnel
      • Telecommuters
  • When
    • The hours of operation must be defined and readily discoverable by end users
    • What sort of support is there during off hours?
    • Who handles alarms and outages?
  • How Long
    • How long should requests take to complete?
      • Completion times based on categories

A helpdesk really needs to have an established policy about dealing with requests concerning technologies that  are out of its scope.  Hard and fast refusals can generate more influential ill will than many helpdesks or their managers seem to care about.  For example, my past experiences with this university's helpdesk with respect to grad students trying to connect their work supplied computers to this network haven't been particularly uplifting.  Fortunately, some individuals have been more willing to give assistance outside of established guidelines.

Defined Processes.  It is often important to at least try to define particular processes for providing help.  Unfortunately, rigidly interpreted but still well defined processes can easily become cumbersome.  Irregardless, there likely needs to be some level of formality in determining things such as

  • who handles what
  • standard ways to handle standard problems
  • standard ways to handle problems
  • ways to handle problems that can be dealt with in standard ways

Even though documenting such things can be tedious and the documentation is too often ignored, it can be very important to develop help oriented documents.

Think of your experiences when calling some sort of online technical support.  Often, it becomes clear the person at the other end of the line is searching internal online sources for information and maybe even step-by-step instructions.  These sorts of resources can help everyone involved.

It can also be important to make serious efforts towards developing a fairly well defined escalation process.  While it may be desirable to have the up front help desk people start most assistance, sometimes they are going to need assistance.  There needs to be someone in particular who can facilitate the referrals appropriately.  Obviously, the better the up front helpdesk personnel can do this themselves the better.

It is almost always essential to have some sort of software to assist the helpdesk personnel and managers track their efforts and determine their performance metrics.  Having places online where end users can go to submit requests in addition to making phone calls or visiting in person is likely to be beneficial. Some other advantages of such an online source are that the data entry is completed by the end user and you may be able to provide some worthwhile sources for self help.

The Icing.  It is often important, especially in largish organizations, to gather more sophisticated statistics about the helpdesk performance.  Some other metrics might include one or some of the following.

  • rate of escalations
  • end user assessed completion rates
  • end user satisfaction
  • services provided
  • etceteras

Providing around the clock coverage can also be an issue for some firms.  Even without such extensive coverage there likely needs to be some sort of call list associated with particular difficulties or outages.

It can also be useful, particularly more likely in largish organizations, to have different desks for clearly different types of problem resolutions.  Who does what must be quite clear to the end users otherwise there will be a fair amount of excess demand just to sort out such issues.  Some firms choose to divide things along lines such as provisioning and initial installation at one desk with problem resolution at another.  One major advantage of such divisions of labor is that they allow for more specialization and depth from the personnel at each location.