Products come and go - the successful ones stay on. I've put out a few notable market failures (the OBOE comes to mind). Not that the circuits are bad, just that it wasn't a machine very many people wanted to buy. The BUGLE PRO was a difficult and clumsy assembly, as it was an afterthought product. When I ran out of FRYLEANER PRO parts I discontinued all machines using that plastic chassis.
Hits and misses. Even Apple came out with a few failures. How many in the following picture can you spot?

My most clever design, the UFO took two years to sell out the ten prototypes. The CASTANET is a superb circuit design, requested by many. It doesn't sell. The 2SK389 JFETs used in the ARCHIVER became too difficult to obtain, so I can't build it (or a FLUGELHORN) anymore. On the other hand, I've personally soldered over 800 HAGUSB. Can't count how many phono PCBs have sold. The BUGLE remains as popular as ever. I've never had to change it.
It's hard to predict the sales success of a product. Good design does not guarantee anything (except low warranty costs). You have to score on an emotional level. Not just sonic performance, but visual beauty, functionality, price, uniqueness, and a lot more.
jh