In the USA, call centres commonly measure Average Speed of Answer (ASA), which is a little easier to understand than NZ's measure of Service Level – the percentage of calls answered in so many seconds, eg: 80/20. An ASA of 10 seconds is easy to comprehend, but damn difficult to achieve day in, day out for call centre managers (CCMs).
The cost of keeping customers waiting longer than 10 seconds rings bells with many CIOs because they usually own the 0800 phone bill. With this in mind, the average speed of answer shouldn't be the sole measure of the contact centre, yet for many it is their measure. Having a balance that includes customer wait times, and managed phone costs would bring more consistency and economic sense rather than just achieving a service measure that evens out the highs and lows.
Just as the call wait times peak and trough, it's really important to understand why an 0800 number bill peaks and troughs, and to what degree, because the peaks may not last long but they can be expensive, as the table opposite outlines.
In New Zealand a relatively large contact centre might have about 65-70 agents on the phones at any time, which means that probably about another 20-30 are vacancies, in training, on leave, sick, or doing anything other than answering the calls. The CCMs find it hard to have as many people on the calls as they actually need; some centres run with just enough, or not quite just enough to reach their service targets.
A headcount freeze
Freephone costs start to escalate when there is not quite enough staff, because the customers' wait time on hold climbs, and this often occurs when there's a headcount freeze on. The telephone bill is amazingly sensitive to the number of agents who are absent when they should be on the queues, or when the centre just needs more staff. In the mythical contact centre example opposite, the impact is very evident. This example centre has 0800 carrier costs of $NZ0.15 per minute, calls that average 240 seconds' handling time (four minutes), and the centre receives about 900 calls an hour.
To achieve the service target requires 67 staff on queue, which will give an ASA of just under 10 seconds. Agents earn $18.00 gross an hour. As the centre is quite large with about 70 agents on the floor, sometimes more as shifts come and go, agents may be unaware just what difference their individual presence on the queues will make.
When the first and second agents log off (the 67th and 66th), the centre barely notices them, because service level only dips below target to 72%, callers are still receiving an ASA of 20 seconds, and if their absence is only as long as the changeover at break time (a couple of minutes), there's little noticeable difference. When the third agent logs off (65th), the impact on the ASA really starts to bite: it drops 10 further seconds to 30 seconds.
Now if the three agents don't return for the next hour, in this centre receiving 900 calls per hour, that extra wait time equates to 153 minutes of additional trunk line time on the phone bill. At 15 cents per minute, that's $22.95 extra phone charges per hour because there are three agents too few on the queues (64 vs 67).
From the table, you can see that the impact of any more than three agents off the queue becomes substantial – ASA starts to plummet, and additional call charges rise dramatically. When the fifth agent (63rd) logs off, not only does the ASA extend out to 87 seconds, but that one person's absence creates an extra $86 per hour in phone charges. It is not uncommon to have to wait two minutes to be answered by a call centre. In the table above, that equates to the situation where between 61-62 agents are on the queues, with the ASA fluctuating between 87-205 seconds. With 61 agents or six fewer than required, the additional costs for the trunk wait time is $266 per hour additional on the phone bill.
The exponential nature of the workload growth shows how the wait times and the call costs just balloon out after a point, and completely dwarf the cost of the missing agents' wage costs. One of the key red flag points in this example is when the 65th agent logs off, because that one agent creates additional phone charges worth well over their own labour cost (1.28 times the hourly rate).
That is the point when the CCM should require staff to go back on the queue from a commercial point of view, let alone a customer service point of view. After that, the added costs just billow.
