My experience is that good investors know that profit+loss projections for certain kinds of products at certain stages are pretty much useless. In these cases,

My experience is that good investors know that profit+loss projections for certain kinds of products at certain stages are pretty much useless. In these cases, your goal is to make something that isn't laughable (when forced to make projections). It's easier to forecast expenses -- you know how many engineers you want, and can guess at legal, rent, etc. But it's also good, if you can't have a market ready piece of hardware within two years, to have an alternate source of revenue in the beginning. For example, you could sell a suped-up (expensive) early-adopter version for a year which will serve as a prototype for the real (cheaper) consumer version. Plan to sell very little in the beginning, as that's what always happens in hardware. And consider planning to sell through someone else. For example, if your device has to do with telephony, assume you will sell through one of the major telephone networks in your area. Then guess at what percent of their users in your target market area you can convert in the first year and calculate your cut of the sale. It's easier to estimate sales if you're thinking about taking a bigger and bigger slice out of a large, defined pie (rather than growing from nothing).