As a PM, for any big project I always try to get a good developer a full two week sprint to do nothing but whatever research he or she feels is appropriate afte

As a PM, for any big project I always try to get a good developer a full two week sprint to do nothing but whatever research he or she feels is appropriate after I have a basic framework of the functionality built out. I'm convinced that 100% of the time, this saves time. The problem is it's tough to quantify the time saved from mistakes or mid-course changes that you would have made without having this time, and it's not easy to get execs to give up two weeks of time from a senior engineer who will really do a good job at this.