Sales Contract Modeling

Hey guys, 

I just got staffed on a manufacturing company that relies heavily on winning a few large sales contracts each year to drive their business. The company sent us their model and pipeline which is structured as a weighted sales pipeline so it allocates probabilities to win each customer. Their actual backlog is relatively small but they have a pretty substantial pipeline of high probability contract wins. However, my boss has established that any contract that is not 100% won and signed shouldn't be considered for our own model. 

I've been tearing my hair out about building revenue projections. Is it really a matter of just approaching it from a bottom-up angle and just build around volume x price? 

Would really appreciate anyone's input on this if they've worked on similar, as I don't have much experience with companies that operate like this.