The reality is very few projects will ever be approved in every field with "I don't know?" as the answer to "How much will it cost?" Estimation is by far the t

The reality is very few projects will ever be approved in every field with "I don't know?" as the answer to "How much will it cost?" Estimation is by far the trickiest project management tool. 2 identical projects with identical caliber teams will come up with very different results. All you can do is use the estimate for a very imprecise first guess to get things started. Of course very little is known for sure early on. The scope isn't final. The tolerated risks aren't final. The project team isn't final. The target quality level isn't final. So what can you do? Admit that these are in flux, and be proactive about managing them. "Ok, now that we know more, it looks like things are more expensive. Do we fix the cost and reduce the scope, or fix the scope to reduce the cost? If hitting the budget is the most important thing, should we preemptively cut scope?" These are the way discussions have to be managed.