- This topic has 6 replies, 1 voice, and was last updated 19 years, 10 months ago by
.
-
Posts
-
I’m doing capacity planning for a call center. One difficulty I encountered is that I only have the average call volume informaiton (e.g., number of calls per day and average call duration) at hand. Is there any rule of thumb or method that I can estimate the peak hour traffic such that I can use Erlang?
There are some industry averages available. What vertical you are talking about.
Regards
Saeed QadriI would be surprised if there was any rule of thumb or method that you could use with any certainty. As with most call centre metrics & processes, there is no industry standard. Your particular problem is dependent on the nature of the business and the set up of the centre (e.g. hours of operation). With the limited data that you have, I would suggest calculating the average hourly calls & AHT for each day and running that through an Erlang calculator. The resulting agent requirement can then be multiplied by the hours of operation and divided by the hours that a typical agent works to give you the base capacity requirement (before shrinkage). Obviously, this doesn’t account for peak hour but will give you a fairly accurate indication of capacity requirements – which should be sufficient to cover for peaks. Alternatively, you could try estimating a % increase above the hourly average….but the only way you are going to be able to solve this problem definitively is to start capturing the info at an interval level.
What is the benchmark for Service Level within Call Centers these days? I run a Call Center with 40 agents and take around 1100 calls per day… what would be the benchmark for Service Level?
Regarding service levels. There is no standard, nor should there be one. Service levels are set to reflect business objectives and customer needs. Each business is different.
As for calculating peak levels from the average, Chris is right. There is no rule of thumb to do that. Some businesses see little difference throughout the day, whereas others have a distinct peak (or two) as well as daily and seasonal changes. Additionally, your ability to tolerate peaks depends on your target service level and agent utilization, which the Erlang calculator factors in the forecast.
If you have to solve this problem “now”, use the average as a starting point and monitor performance and utilization during the day and the week. A good resolution to do that is at least twice every hour. If necessary increase manpower for peak times (remembering that by doing so you decrease the overall utilization). Record call volume for long term trending.
Bill, I guess Joe and Chris have said it all.
I’d just like to amplify whatever Joe and Chris have stated. It is quite difficult to come up with peak hour traffic estimate for any particular account at hand other than the past statistics and experience;especially if you have an Offsite-Offshore BPO model.
Service levels issues are dependant upon the clients and their requirements. If they pay you per-seat-per-hour service level and incomming call load is not the core issue. However, if the client is pedantic about service level then you have to be meticulous about peek hours and all that and in addition to this you need market intelligence from your client.
Saeed, Please send me the pointer to some info on Airline call volumes. I believe that some study was conducted by AAA ?
- The forum ‘Call center’ is closed to new topics and replies.
