Companies can easily fall into the trap of thinking their customer satisfaction is going great simply by looking at how many calls were responded to and the speed at which it happened. Especially if the number of support calls are decreasing. What they may not realize is this measurement might be a bit skewed. What a company really needs to be asking is how many of these calls actually closed an issue for a customer in the way they would have preferred it be handled in today's digital landscape?
Going Back to Statistics 101
Metrics are essentially a way of measuring activity or inventory with objective results. The data is then compared with other measurements to draw conclusions. Statistics can be measured in a number of different ways. Which metric is used and how its used has tremendous value when it comes to help desk activities. Using the example above, simply answering calls in a given time period can produce a metric. If translated, two interpretations can be drawn: the number of calls that were answered, and the rate of those calls answered compared to a previous cycle. This is where things can go sideways. When people get lazy and latch onto a single figure, it can answer a desired management query very quickly, but that doesn't mean the metric really gave good information. We have to go deeper, and this is where correlations really matter. How they are constructed, whether the variables are actually the right ones to use, and whether the correlation is meaningful or just an odd occurrence makes a big difference in decision-making.
What's the Purpose?
Remember, a help desk is designed to provide help. It is targeted to a specific audience and should be providing that service efficiently and effectively. These are mighty words that create tremendous pressure to provide a desired result, but what a company really wants to know is when the help desk is not doing its job. This way changes can be made before more damage occurs. Instead, help desks get handed to the IT department or outsourced to a third party. Both will simplify their performance metric to how fast they can close a call, not whether the call sufficiently answered a customer's question or need. As a result, it's quite possible to get great-looking metrics on the help desk and still have a bunch of really angry customers.
Instead, the metrics used for performance measurement need to be rooted in clearly defined purpose, this way not simply leaving it to the party providing the help desk to figure out definitions.
Defining Good Metrics
The first step is to clearly identify what the help desk should be doing for customers. Is it routing their calls to a subject matter expert? Being an intake center and logging issues for follow-up? Or, is their role actual problem-solving and need satisfaction? The answers to these questions will differ for each company.
Second, the metrics used should be designed on how many cases are completely solved. This metric is often short-changed on how fast problems can be "answered," how many problems were solved in a time cycle, or how many times a customer responded to a feedback survey saying "yes" they were replied to. These are false analytics that don't provide a ton of value. Instead, the measurement should first start with the type of problem. Is it simple information need? Is it a troubleshooting issue? Or, is the customer looking for a repair or refund (i.e. scale problems in terms of levels 1, 2, or 3)? With the categories of possibilities defined, now the actual performance metrics can be applied.
Third, spell out clearly what is a success and what is not for each category of problem-solving. Was missing information found? Good, score that as a problem solved. Was the troubleshooting issue actually resolved? That's a score as well. If not resolved fully, that's a fail. Go back through the steps and work the issue until resolved. Was the customer taken care of with a refund or repair? Score the resolution. If not, then it's a fail. Resist the temptation to solve problems as fast as possible. Doing so can also produce a false metric. The idea is to instead solve problems successfully so the customer is happy and will come back again with future sales and referrals.
Don't plan help desk metrics in a vacuum. Look to those companies who have a really good, working model already and copy them. We don't have libel police when it comes to borrowing good business ideas, especially organizational ones. A number of companies, both online and in real-time platforms have produced really good working systems for the help desk resources. They set the standard for how a help desk should work and the related performance should be measured regularly.
Don't Give Up
A last, but very important aspect of help desk metrics is the need for regular monitoring and compliance to good standards. This is where your performance metrics marry with best practices. Your company should be dedicated to regularly testing. Evaluation of metrics and whether they are on track as measurement tools should be done periodically. If the metric is no longer relevant to the kind of work the help desk is doing, the information won't be of any use either. That will become apparent as the metric results and the purpose of the help desk begin to diverge from each other. But if a company fails to follow up, they will never see this problem develop, and then the whole point of having a help desk begins to deteriorate over time. Good business practices call for regularly checking performance, both of the help desk as well as how it is measured.