What is Average Speed of Answer?
Average Speed of Answer (ASA) is a metric calculating the amount of time it takes to answer a typical call once it has been routed to the contact centre. It is often used with the metric of Average Wait Time (AWT), the period of time before being connected to an advisor.
ASA is often a key part of a contact centre’s Service Level Agreement (SLA), represented as a guarantee to answer a certain volume of total calls within an agreed amount of time. 28 seconds is the global average for ASA, and the 80/20 rule – 80% of calls answered within 20 seconds – is often cited.
In reality, there is so much variety between different centres that there are few widely applicable rules. Research has also shown that customers will tolerate different waiting times for different kinds of service. For example, customers calling technical assistance lines will wait longer for service than customers calling conventional service lines.
Looking for more information on service level, read our article: What Is an Acceptable Contact Centre Waiting Time?
The Formula to Calculate Average Speed of Answer (ASA)
To calculate average speed of answer you need to first divide the total length of customer wait time by the total number of call answered, then multiply this by 100. This is shown in the formula below:
|Average Speed of Answer =||Total Length of Customer Wait Time||X 100|
|Total Number of Calls Answered|
Average Speed of Answer (ASA) and Customer Satisfaction (CSAT)
Customer Satisfaction (CSAT) is a metric often associated with ASA, as time spent queueing is one of the main factors affecting how customers judge the level of service.
The relationship is not binary, however, and low waiting times do not automatically indicate high CSAT. This is, in part, due to how ASA is calculated. The metric is understood from the contact centre’s perspective, starting from when the customer is placed in a queue after they have navigated IVR.
Long or confusing IVR processes can make a customer’s experience both irritating and much more drawn out than the ASA figures would indicate.
Another factor that affects answer speed data is the call abandon rate. If a customer ends the call before they are connected to an agent, their call does not count toward the ASA figure. At the same time, the factor most likely to lead to a customer abandoning a call is a prolonged wait.
The result is that the worst offenders, in terms of answer speed, are often discounted from the understanding of average answer speed.
Want to know more about wait time and CSAT, read our article: Is There a Correlation Between Queue Time and Customer Satisfaction Levels?
Observe the Outliers
The figures can also be distorted by the process of averaging. For example, a contact centre may be open for ten hours a day, adhering to their target ASA of 20 seconds.
However, a closer look at performance hour-on-hour reveals a regular slot – the lunch hour – where a combination of increased traffic and poor scheduling means customers queue for 2 minutes.
This raises the average, but not beyond the centre’s SLA, because throughout the rest of the day, answer times are around 10 seconds. The contact centre is technically on target, but for 10% of the day customers face comparatively long waits. This could potentially do enormous damage to CSAT.
Hour-to-hour averaging reports can mitigate this, as well as a study of call abandon rates for periods of peak traffic. Perspective can also be gained by studying the range of waiting times – the longest and shortest times customers were asked to wait.
Our 2021 survey ‘What Contact Centres Are Doing Right Now‘ found that customer patience appears to be at an all time low, with 43.0% of contact centres reporting that most of their customers will abandon if they’re left in the call queue for only one minute.
If you want to learn a method for predicting call abandon rates, read this article: How to Predict Call Abandon Rates Based on Service Level
For more great information on call wait times and average speed to answer, read these articles next: