I have a predicted volume for 1 month, but no idea on how to arrive at staffing needed. I have a planned AHT as well as shrinkage %. Please help. Thanks.
Question asked by Marisa
Application of Erlang
I’m also curious to hear what others in the industry think. Specifically, on the application of erlang to generating monthly online required hours for long-term staffing/capacity planning.
There seems to be two schools of thought, one being that Erlang is used in the application of schedules but not the monthly requirement, and two that Erlang should be used when generating the online required hours.
My take on the Erlang C vs understanding staffing requirements.
Erlang C will provide accurate staffing requirements for an arrival pattern, SLA’s thresholds, and consider operation of hours.
Overall staffing requirements can be calculate simply, this method is to understand the very basic staffing requirement.
Let’s say my Average Handling time is set for 12 min and 7% planned shrinkage at any given time which considers PTO, VTO, unplanned etc…
I normally calculate on a weekly basis… Lets say 4000 calls a week
12 min handling time * 4000 = 48,000. This says I need just about a half million work minutes that will be working a call.
Consider your staffing requirement. 450 minutes considers 7.5 hours which takes out 2, 15 breaks. This doesn’t consider 30 min lunches because they would not working anyways. So at 100% productivity agents can give 450 min of work effort each. Of course that is not possible and remember we said 7% shrinkage, so let’s take that out.
450*.93 = 418 work min available.
Hmm.. we are not considering realistic productivity.. just because I have 7% shrinkage, do I still expect my staff to work 100% productivity??? Of course not they would burn out quickly.
Productivity is a factor you would have to figure out… In my experience on managing the desk 80% productivity is very good, but see up to 73% most of the time. (Service Desk, not Contact Center btw)
So 418*.73 = 305.14 work min are now available to take and work a call.
Divide 4800/305.14 = 157 FTE / 5 days a week = 31.46 FTE would be required to handle 4000 calls @ 12 min each, with a 7% shrink and 73% productivity.
Of course the % can be changed based on what you think is acceptable for new requirement or based off of existing work.
This only considers the pure workload and doesn’t consider Shift patterns, operation of hours, SLA’s etc…
Think of it this way… To staff a desk 24×7 that would have 100 calls per week, you would need a minimum of 3 FTE per shift, plus .5 or more FTE to cover lunches and breaks, for a 7 day work week.
You would need a minimum of 7-8 FTE to be staffed and available to take that 100 calls a week if you didn’t know when the call arrived or what to expect, that is the staffing required to handle 1 calls or 100 calls a week if you had no idea.
Hope this helps.
Let’s complete the math.
With thanks to snow
Here’s my non-erlang formula for monthly FTE Calculations:
- (Volume * handle time)/3600 = w (workload hours)
- Working Days * Signon Hours Per day = FTE Hours per month
- W / Schedule Efficiency / Occupancy Productivity = Required Hours Monthly
- Required Hours monthly / (1-Shrinkage or Load number) = Total Required Hours
- Total Required Hours / FTE Hours Per month = FTEs per month
You Need to Use An Erlang Calculator
The method posted above assumes that all the calls arrive evenly throughout the day. This is very unlikely.
There is a good article here
You will also need an Erlang Calculator
With thanks to Jonty