"Plan your work and work your plan but don't plan the results."
I've lived by these words and made (and kept) several fortunes by following them.
Can one predict ANY marketing effort? Sure, it's possible. After testing and tweaking, adding data points, capture methods, integrating with reliable Intel, appending data, improving call scripts for call agents, constant metrics analysis and relentless testing.
So many variables.
A friend in my mastermind group tells a great story about how Craftmatic beds used to be sold. Incredible story of phone and in-person selling and conversion. It worked too well until FTC stepped in (probably a good thing).
As to predicting results before you've started, I don't have any idea how you will assign any meaningful number until you DO test.
In direct response marketing there is always a control constant to use as a baseline. It's the tried-and-true methods that, if this formula is used, based on past history, this result will likely occur. Copywriter are challenged to "beat the control."
A great list is better than a great script. Rent the movie Glengarry Glen Ross ;again) and refresh your thoughts about outbound telemarketing and what is most important.
Start with a number, if you must, randomly assigned, as a realistic conversion goal. Run campaigns against that metric and test, test, test.
My challenge: just do something now and ramp up as you get better.