We recently been through this exact same situation. We have a working product, billing, but next to no traffic or users... even initially. We pushed on for a wh

We recently been through this exact same situation. We have a working product, billing, but next to no traffic or users... even initially. We pushed on for a while and tried to cram it down people's throats. When the sales cycle got too demoralizing, we simply took a 1-2 week break, during which we pretended as if we have never worked on our first product. We came up with another product and released it in just 10 days (though by reusing some of our old code), and the reception seems better. We of course hope to some day be able to sell our flagship product as well, in hopes of recouping the sunk cost, but we won't need to count on it anymore -- it would just be an added bonus. In this sense, I believe that it can be wise to: 1) Throw a few (well-thought-out-and-validated) things at the wall and see what sticks. 2) Throw quickly-and-easily-prototypable MVPs at the wall in the first place. Anything that requires 3 months to prototype would present a risk that you probably don't have the stomach to bear. E.g. Twitter probably didn't take long to prototype, same with Formspring. If you can prototype things in 2 weeks, and use the next 2 weeks for marketing, you can pursue 12 ideas in a year, thus increasing the probability of you succeeding.