I wondered if anyone might be able to point me in the right direction on
the following. I'm not entirely fully sure what I need/ and if it even
exists. I'm looking for a software package that might help me achieve some
of the following.

We have a large number of sales representatives, and these all log their
customer calls on an electronic territory management system.
At times clients will provide us with a list of target customers, and wish
to know which of these we can see, and how many times in a given time
period.
I can look at this in a basic way, i.e. either examining what one of our
sales teams has historically achieved on this group, or using all of our
calling data and coming up with different combinations of teams.
What I am struggling to do is be able to say what a new team for example
could do based on the historical calling. For example, I could see how many
times on average each customer has been seen, but then need to model this
back to what is possible for one team. I know any particular team of a
given size will achieve seeing about 40% of their customer base in a year,
yet all teams combined achieve about 80%. I need to know which of the 80%
one particular team is more likely to see.
I can supply any model with the ease with which each customer can be seen
and other metrics.
The bit I'm finding really hard to determine is the following- imagine
there are three customers, each can be seen as easily as the others, and
you have seen two of them once, and the other not at all. What is more
likely, to see the third one for the first time, of one of the others for
the second time.

Is there some software or technique for analysing past calling patterns and
using these to predict likely future calling patterns, based on set
objectives, targeted customers etc?

Thanks

Graham