My suggestion is to do your marketing up front, concurrent with development, to get incremental buy-in and iterate concepts until you find a good one, instead o

My suggestion is to do your marketing up front, concurrent with development, to get incremental buy-in and iterate concepts until you find a good one, instead of on the back-end. I find you have much better results that way. You can arrive at something that 'sells its self' by aggressive customer contact and pre-selling. But in fact, it didn't 'sell itself,' you just did your marketing up front to make sure you built something people wanted to pay you for. As to marketing being able to sell bad products, I reply: New Coke. ;)