Run them like hell
Run the numbers, please.
Run them as soon as possible, and as deep as you can.
When the projections are performing like an equilibrist between profits and loss, it’s safe to assume shit will happen. So you better fix the business model or you call it off.
On the other hand, great numbers can mean one of two things: you have some huge potential (and should proceed with caution so you don’t blow it) or you are fooling yourself with wrong assumptions.
Happy numbers must be put under scrutiny over and over and over.
And of course, revisit them as soon as you have new pieces of information that may modify your previous assumptions and projections.