Sometimes you have an idea, and want to have a pilot or proof-of-concept built, either to prove the viability of the idea, to show it to investors, or to estimate the development costs of the "full" project.
Besides being a programmer, I'm also a businessman, so I grasp business ideas quickly, and turn them into working software fast.
See the case study on how much can be learned from a prototype.
I also offer advice on which technologies to use and which to avoid, from both a technical and business standpoint. (For example, I would advise against using Java for a simple web project; but would recommend it if you're a bank looking to build a new credit-rating engine whose complexity will be duplicated every year for the foreseeable future.)