Shelf Awareness of June 28, 2019 included a column by Robert Gray under this title, now archived at his site, Fresh Eyes Now.

Mr Gray has been reading H. A. Pavey’s 1905 piece from the Chicago Daily Tribune exploring the question “Why Novel is a Success” [sic].

Mr Pavey basically ascribes the fog of uncertainty surrounding this question to the illogicality of women who make up the majority of buyers of novels! (He was writing in 1905.) He goes on, quietly changing to the masculine pronoun, describing a reader going into a bookshop and picking up a novel:  “Instinctively he opens it at the first touch. Type and paper will be expected to make the first appeal in the physical makeup. An attractive frontispiece and title page will be convincing, as will possibly well done illustrations. Then the scrutiny of the cover will follow.” This is all very flattering to those of us who have toiled on the physical side of the book business, but unfortunately it isn’t enough to seal Mr Pavey’s deal. “In the meantime the salesmanship of the salesman will be called upon as it so seldom is at the average department store’s general counters. For any book that is in demand, the salesman will have had his own brief lesson. He will have read the reviews of the book as far as possible; he will have run through it himself perhaps as closely as does the average reviewer; he has at his tongue’s end a striking situation or two of the situations needed to have made the work talked about and favorably reviewed.”

However Mr Gray concludes, along with Mr Pavey, that the reason for buying a book is different in each instance, and impossible to discover. “So, what’s the magic key to discovering why readers buy particular novels? . . . there is none.” If only we did know why people buy books we’d be in a much easier business: we could promote the books directly to the people who we know want them, and as a result we’d be able to judge ahead of time how many copies we ought to print in order to fulfill demand. (Certain types of academic publishing already approach this condition.)

Ultimately the ability to fulfill exact demand by print-on-demand will bring us close to this situation. This doesn’t mean that we’ll do it though. The temptation to get a lower unit cost of production (to increase your profit margin at a given retail price) will make publishers continue to gamble by filling warehouses with speculative stock. The best we can anticipate, at least until the book manufacturing industry withers away, is that we’ll use POD to fill the inventory needs of that famous long tail — the last few copies which dribble out over the years when the book has substantially been forgotten. Being able to fill demand exactly won’t, of course, tell us why this or that novel is a success, but as long as we can sell as many copies as possible publishers will perhaps not be altogether concerned with retail customer motives.