This is a tricky one. It is best to build a bottom up forecast. You need to internally decide your sales strategy and make your best estimate of how many customers you will sell, how much you will sell to them and how much repeat business you expect to have. You will do this for each market you will sell to. I would be as granular as possible when building this internal forecast. If you have specific target customers that is great include them. Combine all of this for the next 3 years and you have your forecast. It will be wrong by the way, but it's all you have to go on.

The real problem is when present this to investors. There is a tendency to give too many details and for some people to get caught up in the details. I would provide potential investors only the high level projections and explain the process you used to derive the numbers. Once you dive down into the details of a plan like this you are inviting all sorts of criticism that you really cannot defend, then it can turn into a game of gotcha.

Any good investor will know that the projections are wrong but should appreciate the process that you have used to create the projections. At the end of the day they will either think that what you have created is a good representations of your product/service potential or not. They will not benefit from getting too caught up in the detail and you certainly will not. Keep it simple and try to steer any conversations back up to the substance of the projection rather than the details.