Low-cost, High-Performance IT Helpdesks

Updated: March 24, 2010

IT helpdesks, or IT Service Desks if you're going ITIL, have long been stereotypically perceived as low-status, temp-job environments and money sinks of the necessary-evil type for businesses. A large organisation can easily spend six to seven figures a year, or more, in maintaining such groups.

And yet, when put under the microscope, there is no reason for them to be this way other than "everybody knows that's what happens".

Every major helpdesk I've encountered (including the outsourced services) has had this problem. And it's not just an issue with image. It translates directly into effects on the bottom line. But money continues to be poured into underperforming systems because of two reasons. Firstly, no-one wants to be assigned the relatively low-status job of auditing the IT helpdesk. And secondly, hardly anyone actually knows what the heck the junior propellerheads are doing in that room anyway, so auditing is a pain and it's really difficult to set realistic goals of improvement and efficiency gains without constantly running into things which can't be done because the employer has the wrong hardware, the wrong software, the wrong training, the wrong employees, or the wrong policies.

So employers either put up with it and perpetually grumble about the cost of a team they can't see the value of, or they outsource to a third party who charges twice as much for half the service but produces very nice monthly pie charts and can be used as an emergency scapegoat.

Meanwhile, the few groups of people who actually are in a position to analyze and improve the service are hamstrung. The technicians themselves are mostly junior short-termers, bucking for a job higher up the IT structure and being hammered with constant phone calls and emails. When they're not dealing with endless requests for password resets and issues which can be solved by turning the computer off for three seconds, they're swamped in the administrative record-keeping requirements of their ticketing systems or training their own replacements. The few long-term staff tend to keep their heads down and never question anything. And the managers of these teams find that much of their time is spent running around stamping out brushfires because of the constant high churn levels and minimal experience levels of their staff. In addition, few helpdesk managers have the personal experience on the phones necessary to root out inefficient processes and procedures which are being passed down from employee to employee. Their budgets are drained by constant training, retraining, requirements of new staff, trainees who aren't up to speed, and the need for more supervisor/nanny positions than most other business teams.

But it doesn't have to be this way.

It's entirely possible to transition quickly and cleanly to a low-cost, high-output IT helpdesk where the technicians are long-term, knowledgeable, and helpful, requests are processed quickly and efficiently, the team is a much tighter firewall for the higher-level IT groups, and a far greater proportion of requests are handled at the first point of contact. A team where churn is low, training is low-cost but highly effective and fast, new staff are up to speed rapidly, and the team itself generates summarized real-time business data with values far in excess of the cost of running the group itself. In short, it's entirely possible to have a great service with fast turnaround which is actually generating business value on a daily basis.

How?

Very easily. While the details for each implementation will be different depending on the business structure of the employer itself, several central tenets stand out.

  1. Move non-IT functions away from the helpdesk.

    This may seem to be obvious, but the IT helpdesk in any large organisation tends to wind up as the dumpster for anything that other, more high-status areas don't feel like doing themselves. As an example, if you have 400 employees for every helpdesk phone, and the lines are jammed every Monday morning, consider that resetting a password is _not_ an IT issue, it's a security issue - and it's best handled through an employee's direct chain of command. And who are the regular staff calling when they need computer training?

  2. Move mass-effect functions away from the helpdesk as the business scales up.

    Having helpdesk staff walk around to an employee's desk to press the reset button on their computer might make sense when there's only five people in your entire staff. When you have five thousand, scattered across multiple buildings (and sometimes multiple cities), that's not really plausible any more. Either train all employees to perform basic IT troubleshooting (a card attached to the side of each computer is a start), or have a designated person in each office or on each floor who can perform perfunctory assistance such as rebooting a PC, checking all the cables are plugged in properly, and knows the fixes for half a dozen common faults, failures, and error messages.

    Train staff to go to this person first, and to have that person be the one to call the helpdesk if things are really broken. When you have dozens of offices, it's a lot more effective if the helpdesk receives only a dozen "the network is down" calls rather than five thousand. In addition, it can be a lot faster to have a person five desks away look directly at a problem than someone five states or countries away having a "best guess".

  3. Reduce churn costs, training costs, initial trainee slowness and ignorance costs, stress, team size, employee absence costs, and costs-per-ticket-processed in one fell swoop just by paying your helpdesk technicians above-average salaries. You can attract top-level technicians for a salary less than a third of what you'd pay the equivalent number of average-quality staff. And there are methods and policies which enable this imbalance to be pushed even further - it's quite possible to get it down to a quarter or even a fifth of what it was costing previously.

  4. Reduce, revise, or remove administrative processes which are slowing down ticket production and calls. A technician who doesn't have to spend half their time ticking boxes, selecting dropdowns and switching screens on a slow ticketing system is a technician who can take twice as many calls per day and will be a lot happier about it.

  5. If you have an in-house programming or IT development team, there are enormous leaps of efficiency improvement across the board which are possible by taking into account data which the helpdesk team already collects automatically.

  6. Is your CIO or IT director getting real-time updates on the statuses and capacity issues of major systems? The helpdesk is quite often the first section of IT to know about such things - often even before built-in monitoring systems have alerted anyone. Are the upper levels of IT and management plugged into this resource?