Believe it or not, one of the first tips given to me during my freshman year at college when I started working at the campus help desk was fake it till you make it.
The idea was, it took some time and effort to understand all the various pieces of information that made up the “answers” and “solutions” of the help desk. In the meanwhile, there was a job to be done and calls to be answered. So it was best to grunt a few yes’s, ask a few “Lets try rebooting your computer” and “well, it’s a hardware issue, you will need to bring it in”.
Then, when you eventually learnt the right answers, you could stop generalizing and get more actual help-ing done.
Over at Ask the VC, someone asks How do I create Projections for something that I don’t know or am not sure about right now. The Projections are needed today, but I won’t have the information for a few more weeks or months. Whats to be done?
This is a very standard problem at startups. Should I bother with making the damn product or actually spend time guess-timating cost and revenue figures. Without a beta product, any guesses are all going to be useless, right? Wrong! Might be, faking it till you make it isn’t the right solution for such problems.
Or, to be more precise, get your equations thought out logically and use variables (3a + 2^b = cost of deployment, for example). Then, even if you plug in wrong numbers, all you have to do in the future is just change the value of the variables. You always know which side is up, you just need to adjust your altitude!
Sure you could have messed up your initial assumptions, but just getting to outlining the various costs/metrics will help you “fake it” better.