Hanging on the telephone: some practical considerations for crisis communications

Introduction
Most organizations are now fully aware of the critical importance of crisis communications in their business crisis communications planning: business crisis communications journals, magazines and conference schedules are often dominated by discussion of crisis communication issues. However, most of the effort and resource tends to be directed toward outgoing communications, particularly communications with the media.

This often results in a lack of consideration of an equally important issue: that of handling incoming communications. In the aftermath of a disruption, organizations may have to handle large volumes of telephone calls from a variety of sources including:

Staff and families;
Customers;
Suppliers; and, of course,
Media organizations.

Whilst these are highlighted as issues to consider in almost all business crisis communications plans, there is generally little mention of the resources required and how these will be organized. This article illustrates why dealing with large numbers of telephone calls in a crisis communications situation requires detailed planning in advance and aims to provide some practical guidance for call handling.

The model
The graphs used throughout this article are all derived from a simple queuing model constructed in Microsoft Excel. Given an average call volume, the model simulates calls arriving at random time intervals. If an operator is available when a call arrives the call is answered immediately, otherwise callers have to queue until an operator is available. This time between a call arriving and it being answered, the ‘waiting time’, is used as the principal measure of the quality of service that callers experience. In addition the model makes one simplifying assumption: namely callers are prepared to hold on indefinitely (i.e. no caller rings off regardless of how long they have to wait for their call to be answered).

A simple scenario
Typically the handling of incoming calls is treated as a functional responsibility i.e. HR staff handle calls from staff and families whilst corporate communications staff deal with media enquiries. This policy results in a number of small, separate teams so at this stage we will assume that there are three telephone operators on duty at any time in a team. We illustrate the use of the model with the following scenario:

The team receives an average of 20 calls per hour; and

Each call lasts between 2 and 10 minutes (with an average call length of 6 minutes).

This scenario gives rise to the following distribution of waiting times:

The first important observation from using the model is that, because of the random nature of the process, we can never guarantee a maximum waiting time. However, the model indicates that, under these circumstances, we can expect:

58 percent of calls to be answered within 1 minute;
78 percent within 3 minutes;
91 percent within 6 minutes; and
99 percent within 12 minutes.

Is this a good or bad performance? What is a reasonable length of time to expect people to wait? As mentioned above, calls will come from a wide range of sources: is the reasonable waiting time the same for all groups? Defining an appropriate service level (for each group) is therefore the first challenge. Inevitably this must be defined in terms of percentages of calls being answered within a certain time rather than an absolute maximum waiting time.

Call volumes

Obviously, as the volume of calls arriving (or the duration of calls) increases, waiting times ramp up. The graph below plots performance, against two (arbitrary) service level benchmarks, as the volume of calls being received by the team of 3 operators rises from 20 to 28 an hour:
Percentage of calls being answered within 3 minutes; and
Percentage of calls being answered within 6 minutes.

Clearly the performance against both service level benchmarks degrades rapidly. Furthermore, as shown in the graph below, at a volume of 28 calls per hour, a significant number of callers are waiting for an extremely long time as only 55 percent of calls are being answered within 12 minutes.

The system is clearly very sensitive to slight changes in call volume (or call duration). A rise of only 40 percent in the number of calls has led to a seven-fold increase in the number of callers waiting more than 6 minutes and a 45-fold increase in the number waiting more than 12 minutes. Worse still, if call volume exceeds the theoretical capacity of 30 calls an hour, the operators can no longer cope and waiting times increase indefinitely. It is therefore vital to have some realistic estimates of what call volumes to expect in order to allocate appropriate resources.

Pooling resources

Whilst there are obvious advantages in routing calls straight to appropriate subject matter experts, it is clear from the previous section that small, independent teams can rapidly become overwhelmed by even a modest fluctuation in call volume. If instead, resources are pooled to form a single call-handling team the system becomes much more robust. The graph below shows the result of combining 4 teams of 3 operators, each receiving 28 calls per hour into a single team of 12 operators handling 112 calls per hour.


Performance against all the criteria discussed above is significantly improved:
Percentage of calls answered within 3 minutes is up from 20 percent to 51 percent;
Percentage of calls answered within 6 minutes is up from 33 percent to 77 percent; and
Percentage of calls answered within 12 minutes is up from 55 percent to 95 percent.

Putting theory into practice

This article has highlighted some of the issues that need to be considered in preparing a crisis communications plan. It is clear that some thought needs to be given to what service level is required and, where possible, this should be discussed with stakeholder groups to ensure that no false expectations are created. It is also important to make a sensible estimate of what volume of calls can reasonably be expected; this will obviously vary greatly depending on the scenario. Based on these two factors one can then decide how much resource is required.

Having calculated the overall resources, two important decisions have to be made:

Do we handle telephone calls ourselves or outsource; and

If the answer is the former, is call handling organized in small functional teams or as a single integrated team?

The arguments for and against outsourcing are familiar from other aspects of business crisis communications management and, indeed, business in general. The second question involves a trade-off between the increased cost of cross-training telephone operators to handle a wider range of enquiries ahead of the event and the much improved flexibility this provides during an incident.

More importantly though, this article has hopefully illustrated that handling large volumes of calls in real time is a complex issue and should be avoided if at all possible. The most important message is, therefore, to communicate effectively with stakeholders as part of the crisis communications planning process: to prepare them so that during any disruption they access alternative sources of information such as recorded messages and websites. If the overall number of incoming calls can be reduced in this way the available resources can be focused where they are really needed.

Source: Patrick Roberts link