Many contact centres seek to improve customer service. Achieving it can be a challenge. The first step is to agree on the measures which identify acceptable levels of customer service. These will vary by company but are likely to include one or more of the following:
- Average call waiting time
- Average email response time
- Grade of service – e.g. the number of calls answered within a defined period
- Average handling time (AHT)
- Occupancy – the percentage of time agents are speaking "live"
- First call resolution (FCR) rates
The next step is to have a clear view of forecast traffic volumes across all channels, ideally down to fifteen minute intervals throughout the day. The aim here is to identify peaks and troughs in activity so that contact centres can ensure they have the required level of agents available to meet demand. However, traffic forecasting has its own set of challenges.
The difficulties of traffic forecasting
The start point for forecasting traffic volumes is historical data drawn from the automatic contact distributor (ACD) management information system. This should provide call patterns by hour, day and month. However, these days most contact centres handle more than just voice calls so similar data needs to be captured for all channels, including email, SMS and chat.
This data will only be historical in nature, which is a useful start, but will contain anomalies. For example, interaction volumes may have spiked in one month as a result of some marketing activity. Understanding the impact of such events will be crucial in order to prepare more accurate traffic forecasts in the future. It will then be possible to build in contingencies to cater for future known events.
However, actual interaction volumes can be influenced by factors that cannot be foreseen. Consequently, provision should be made to respond to overall traffic levels as they occur. This means being able to track traffic volumes in real-time.
Matching agent availability to demand
Contact centre managers need to ensure that sufficient numbers of agents, with the required level of skill and experience, can be called upon as traffic volumes fluctuate during the day. Planners must take into account events such as holidays, sickness, training, internal meetings and preferred shift patterns.
Having identified forecast traffic volumes, determined service level standards, and compiled a list of agent availability, it should be relatively easy to match demand to supply. The problem is that both demand and agent availability will vary each day. Adjustments need to be made regularly, perhaps a number of times throughout the day.
Making such changes manually is time consuming and prone to human error. That is why contact centres adopt workforce management (WFM) tools. Solutions such as QMax™ take data from the ACD information system and automatically compile highly detailed and sophisticated traffic forecasts. These can be adjusted to reflect known events and can also amend forecasts using 'what if' scenarios.
WFM tools also accommodate agent availability as it fluctuates on a daily basis. However, what is perhaps most impressive with proven WFM tools is the ability to react to real-time traffic volumes and automatically adjust agent schedules to match the volatility of interactions. It means key performance metrics reflect the contact centre's ability to meet the demands of its customers. This illustrates how WFM tools can improve customer service and the business case is compelling too. WFM benefits virtually all contact centres with thirty seats or more. Operating costs can be reduced by up to 10% and payback can be under twelve months.