ITIL says what needs to be done (requests for IT support should be taken at a single point of contact) and why it needs to be done (it allows the users to perceive a single front for IT and provides for more consistent information).

ITIL does not say who (the Service Desk could be staffed by the CIO and their direct reports for all ITIL cares) or how (ITIL has no recommendations for the calls being answered within x seconds with an average call length of x minutes).

That being said – I can say from many years experience manning and managing a large Service Desk (700+ calls a day) that a good tier 1 support person can handle between 40 to 60 calls per day. This is entirely dependent on:

-The complexity of the environment
-The nature of the calls (majority password resets, application issues, hardware issues, etc)
-The tools available (access to all security systems, remote control software, accurate inventory)
-The skill level of the analyst

I set the record at the Service Desk just to see how many calls I could take. It was the first day back after a two week company wide shut down for the winter holiday, 99% of all the calls were password resets, I type at 120 wpm, I implemented the Incident Management tool and knew how to log tickets with the fewest keystrokes, and I almost never touch the mouse. I logged 217 calls in 8 hours.

There was a mathematician born in the late 19th century who came up with a calculation for how many trunk lines would be needed to handle a projected volume of phone calls. The calculation is named after him and there are three varieties. The Erlang A calculation is for trunk lines for non-queued call (callers get a busy signal if all trunk lines are in use), the Erlang B calculation is for trunks where calls can be held in queue, and Erlang C is for number of operators needed to handle a predicted number of calls.

The Erlang C calculation is perfect for calculating the staffing of a Service Desk if you have the parameters of:

# of calls per hour
max desirable queue time
average in-call time
average after-call time

Do a search for Erlang C on Google and you can find lots of calculators.