It might be just me but I think a lot of people read "Selling ideas" as "demonstrating the technical merit of ideas". Unless the requirements are brain-dead si

It might be just me but I think a lot of people read "Selling ideas" as "demonstrating the technical merit of ideas". Unless the requirements are brain-dead simple, you'll need to sell at the very least the criteria on which you're measuring technical merit: is a latency number a hard requirement, will it scale for x of years, will it be buildable in y months or years. At least some of that is going to come to convincing someone you've done your homework even if they couldn't make a better decision right?