Brian Hibbs talks about discoveries made due to the accuracy of his POS system, and discovered in four months, lots of graphic novels didn’t sell a single copy:
What kind of flat-out shocked me is that something approaching half of the trade paperbacks we carry have not sold a single copy since we’ve installed the system. Now, admittedly, I’ve been telling myself that my goal was “two turns a year” as a minimum for anything to stay on my rack, and we’ve only got four months of solid hard data (so, two months to go!), but I’m really starting to see just how big the anchor is that’s slowing us down. Come February, I’m probably going to get extremely ruthless about what to start cutting away before we drown in the sea of SKUs being unleashed upon us.
The main categories?
1) “indy” books that have neither any significant word-of-mouth, known or established creator, or clear and specific concept which “sells itself” to the customers.
2) Mainstream Superhero B-list-or-less material.
Hibbs has much more in the link about both these categories. See also: Johanna discusses Hibbs’ column and another store which gave up selling back issue comics.