To attain a top performing service desk, it is important to implement and leverage metrics. As in other IT areas, measure what is important, not what is easy to measure. Ensure that you have a clear focus on the outcomes that you are working to achieve and that your metrics measure your progress towards those outcomes. Your primary metrics should be inherent operational measures and ‘outcome’ focused. They will essentially measure whether you have or have not delivered what is required of your service (e.g., did you pick up the call quickly enough, did you resolve enough calls at the first point of contact, what is your customer satisfaction score). Augment these primary measures with performance metrics are those that provide you with internal insight into how you improve the performance of your service (average call duration, standard deviation of call duration from the mean average, number of short calls, average calls per agent, average contacts per supported user, incidents / requests fulfilled within SLA).
- Targets and measures can have an ‘equal and opposite’ effect and this needs to be managed closely. The purpose of setting a target is to influence behaviour to achieve a desired outcome. The purpose of measurement is to gather facts through which we learn information and then respond accordingly, changing behaviours as required. Poorly set targets can generate negative behaviours in both agents and leaders and it is important that consideration is given to the ‘down side’ in both the design of metrics and in the policing of them once in operational use. For example, setting the dominant target as call length can result in agents rushing to close the call without fully resolving a customer’s issue or delivering the service.
- A good filter to assess the quality of your metrics is to review which of them do (or should) drive a change in behaviour. If a metric can ‘go red’ without prompting action it is likely to be meaningless. Those metrics that measure the key outcomes for your department are the basis of your control framework and a healthy mix of lead and lag indicators should be developed to ensure that there are no surprises with issues picked up early and preventative rather than remedial actions deployed to correct the situation (for example, run your telephony numbers each day with a cumulative total month to date and a forecast prediction through to month end).
- Where possible, measure from the customer in, rather than from your agents out when setting service targets. Best practice service desk metrics include:
- The average time a customer waits for their call to be answered.
- The % of customer calls presented to the Service Desk that are answered by the Service Desk (conversely the % of customers that abandon prior to being answered).
- The % of customer calls answered by the Service Desk within an agreed time threshold (here, % answered within 20 seconds).
- The % of customer calls that are resolved by the Service Desk at the first point of contact.
- Customer satisfaction with the Service Desk service.
Metrics appearing reasonable on paper don’t always translate into a high level of customer satisfaction (much to the confusion of both the Service Desk and the business). If we take two of the above targets and use some fairly typical industry standards to measure the customer experience, we can observe the following:
Classify a ‘successful’ call as one in which the customer has their call answered within a timely manner (20 seconds or less) and their issue is resolved by the agent on the call. If our targets are 80% answered in 20 seconds and 70% resolved on the call, assuming that we are not under or over performing, the % target also represents the probability of that event occurring whenever an individual call is presented.
0.8 x 0.7 = 0.56
When our Service Desk is delivering against these two operational metrics, the probability of any individual call being classified as ‘successful’ is therefore only 56%. That is nearly 1 out of every 2 calls will not have been ‘successful’ in the way that we have defined it. Whilst we said that we would answer only 80% of calls in 20 seconds and fix only 70% during the first call; that often isn’t the customer expectation – they expect to be part of the 80% and the 70% and when they aren’t, they are dissatisfied. Thus, while we have set two reasonable targets in isolation yet together they do not yield the customer experience we appear to promise.
When agreeing your targets with your clients and customers, ensure that you are realistically representing what the customer experience will be like and use measures that set those expectations accordingly. For example, rather than a goal to answer 80% of calls in 20 seconds, answer 100% of them in less than one minute. This sets an expectation with every caller that they may have to queue, but if you design your service properly, they will not queue for beyond a small but reasonable amount. Further, work to drive the call resolution rate higher (75 or 80%) as well as the customer satisfaction level with the handling and effectiveness of the agent. This would then yield a better experience and overall result the customer and meet their expectations. Being open and honest with your customer when setting targets may present some challenges, but it is much better than spending the rest of your relationship together justifying an expectation gap. And your team will rally much more around realistic targets that also demonstrate real improvement in the customer experience when they deliver better.
Establishing the right outcome based key targets and metrics is essential to the effective and efficient running of the Service Desk – if we haven’t been clear about what great looks like then we could only ever get there by accident and wouldn’t recognise it when we did. Metrics should drive fact-based decisions on how to run operations and improve. If the information does not inform and drive behaviour, then it is not the correct information. Achieving metrics is not a meaningful purpose in itself and it is too easy for the numbers to become more important than the reasons we created them in the first place. We need to persistently ensure that our teams stay focused on the customer outcome and not just the number.