Looking for Expected Wait Time tips

[Milesprower]

I apologize in advance as I know there has been a lot of good info on Expected Wait Time (EWT) in this forum, but the search function for the Symposium forum has not been working for me for some time (I have sent messages to Tek-Tips, but have not received a response).

I have a customer with small to large skillsets (10-30 agents) that wants to implement delay in queue announcements based on EWT. They received feedback from Nortel to try Average Speed of Answer (ASA) when the number of agents/call traffic is low instead of EWT.

So, here's my question(s): Has anyone used ASA for small skillsets? What would be the cutoff (number of agents/call activity)? When would you switch over to EWT (call traffic/ agents logged in threshold)?

I'm thinking the logic would be something like this:

ASSIGN 0 TO delay_cv
/* choose delay value based on agent/call traffic */
IF LOGGED AGENT COUNT < agent_threshold_count_gv OR
QUEUED CALL COUNT skillset < queue_threshold_count_gv
THEN
ASSIGN AVERAGE SPEED ANSWER skillset TO delay_cv
ELSE
ASSIGN EXPECTED WAIT TIME skillset TO delay_cv
END IF

WHERE delay_cv EQUALS
VALUE 0..60: GIVE RAN short_ann_gv
VALUE 61..180: GIVE RAN med_ann_gv
VALUE 181..600: GIVE RAN long_ann_gv
DEFAULT: GIVE RAN long_ann_gv
ENE WHERE

 

[captaingadget]

I have to admit I'm a little puzzled why anyone would want to switch from EWT to ASA for low traffic reasons. Are Nortel suggesting that EWT is inaccurate because I cant understand why this might be.

ASA by it's very nature is an average based on historical data. EWT is based on a combination of real time and historical data and takes into account the position of the call in the queue, number of agents and average call duration.

What is absolutely vital is that the call is queued to skillset before any EWT or ASA intrinsic is executed.

 

[Milesprower]

EWT is inaccurate if you have small skillsets or spikey traffic (but as you say, ASA suffers from the same issues). A very seasoned Nortel engineer happened to be working on a different issue when the customer expressed an interest in EWT. As the suggestion came from Nortel, the customer wants to follow the recommendation.

 

[captaingadget]

Just read the scripting guide, not a great help really. It says that the ASA is based on the 10 minute moving window real time data, and that it is an average of a call entering the system with a given priority. Logic suggests that if they haven't had a call for over 10 mins then there's no data.

EWT is based on real time and historical data and is specific to an individual call. I guess you need to try both and have a play to see whats best.

A very interesting question Miles, we could debate this for hours I think.

 

[Milesprower]

You're right. I was hoping to not have to debate the issue with the customer (for hours on end).