An 'abandoned call' is created by a caller, hanging up, before their call is answered - also called 'lost' or 'missed' calls.

It could equally be argued, an 'abandoned call' is created by a call centre, having insufficient agents to answer calls or having too many incoming trunks (lines), unmanned.  As a 'rule of thumb' the service level headcount requirement multiplied by 1.5 gives the required number of trunks.

Calls can be abandoned during IVR messaging (whilst listening to queue selection messages), during ringing (queuing), on hold (agent may be seeking advice), ringing at agent turret (agent may be in an available state, but may have walked away from their desk). The generally accepted definition is where calls are abandoned during ringing (queuing).

Abandoned calls are a key component of Calls Received (Calls Answered + Calls Abandoned = Calls Received).  The Percent Abandoned formula is Total Calls Abandoned / Total Calls Received.  Some call centres use Answered instead of Received - this is incorrect and provides a more favourable percent.

'Short Abandons' are not often reported by companies.  These are abandons, which occur during the first 5/10 seconds of a call.  The reason being is that the caller may have misdialled and therefore, hangs up when the mistake is realised.

Automatic Call Distributors (ACD) can generally be programmed to exclude 'short abandons'.

Abandoned Calls, whilst a key call centre metric, must not be targeted alone - Abandoned Calls are the result of not answering calls quick enough - but should form part of a suite of metrics - Service Level is the key metric, with indicators being Abandoned Calls and Average Response Time.

Some call centres target only Abandoned Calls and this is full of problems:

• ​It is not possible to calculate headcount resourcing from Abandoned Calls.
• It is not possible to determine the customer experience.
• The level of customer wait times will be unknown.

A monopoly utility is a good example of the above.  Company x targeted only abandoned calls, which were maintained at a level of 10% over 3 years.  They didn't believe that it's marketing or new payment procedure was having any adverse effect on customers as abandoned call levels were not rising.  However, they were getting further and further away from their customers.  Company x was not monitoring Average Response Time, which over the 3 year period had increased from 35 seconds to 180 seconds.  As customers were in a monopoly situation, they became accustomed to poor service and would 'hang-on' once they got through.  The result of this is that Abandoned Call levels were maintained but Average Response Time increased.

Some call centres believe they can select metrics off different 'shelves' - company y decides on the following selection:

• Service Level 90% answered within 15 seconds
• Abandoned Calls at 10%
• Average Response Time at 35 seconds

It cannot achieve it's targets as they are not related - Abandoned Calls and Average Response Time are corollaries of Service Level.

A Service Level of 90% within 15 seconds will likely yield the following;

• Abandoned Calls < 5%
• Average Response Time < 5 seconds