If a once-remarkable product is no longer remarkable and is no longer capable of being perceived as remarkable, Seth Godin tells us, then you'd be more successful by abandoning it, and working on something new, something remarkable.
Purple Cow - Transform Your Business by Being Remarkable," Godin discusses what Procter & Gamble (P&G) should do with Tide, their famous laundry detergent. From my perspective, it would be crazy to abandon a money-maker like Tide. The folks at P&G seem to agree with me. They're now offering Tide Pods, Tide with ActiLift, Tide Vivid White - Bright, Tide plus Febreze, Tide plus Febreze Sport, Tide Plus Downy, and Tide HE with ActiLift. They also offer Tide Free, a product whose packaging is even free of the orange color that millions of people associate with Tide.
microISVs with software applications that were once remarkable should try to find a way to make them remarkable again. And if that's not possible, I think you should continue to market them for many years to come.
Godin says that if you introduce a new product, most people won't buy it, even though it is indeed remarkable. Most people are happy with the old stuff that they have already. For the software developer with a new, remarkable application, this can be frustrating. But for developers whose products are established and bringing in significant revenue, this is good news. It means that you have some time to bring your software back into its former leadership position.
My recommendation is to set your sights on creating remarkable software. But if you end up with a solid, unremarkable application that brings in respectable revenue every month, then continue marketing it and go for "remarkable" on your next development project.