Let's look at the publishing starts of the Twilight and HP series, then.
Twilight's editor claims to have recognized the book's potential halfway through the MS. After a 2003 auction, Little Brown upped their $300K offer, and Meyer received an advance of $750K for three books. Initial reviews were strong, and Twilight debuted at #5 on the NYT list within a month of publication.
So this is a book that started out with high expectations (and back-up) and fulfilled them.
Whereas Bloomsbury paid only 2500 pounds for HP and the Philosopher's Stone, then did an initial printing of only 500 copies. 300 of these were distributed to libraries -- in other words, the editor was betting on a word-of-mouth campaign. Through 1997 and 1998, HP/PS garnered positive reviews and a number of UK awards -- most tellingly the Smarties Book Prize, which is voted by children. In other words, the readers themselves.
The American edition (Sorcerer's Stone, because everyone knows that Americans are stupider than Brits and will recoil in horror from the word "Philosopher") sold for $105K (very decent for a kid's book at the time.) It was published in October 1998, started garnering good reviews and awards, and finally topped the NYT list in August 1999. Ungraciously, it then parked at or near the top until pressure from (nonScholastic) publishers urged the Times to create a separate list for children's books.
So, much lower expectations for HP and a slower build-up, but one that led to a juggernaut -- I'd say on the merits of the book, which appear to have hit a high percentage of the readers. Or, as an earlier reviewer in The Herald noted, "I have yet to find a child who can put it down."