Library acquisitions lore contains a cautionary tale of a patron in a demand-driven environment who spent a huge chunk of the library budget on ebooks about bananas. This story and others like it have been used to perpetuate the argument that demand-driven acquisition will result in collections that don't appeal to a broad audience or are otherwise unbalanced. We apply post-acquisition usage data from multiple libraries to test the hypothesis that patron-acquired versus librarian-acquired collections have different usage profiles. In addition, we analyze their subject profiles to evaluate collection breadth and balance. Our results will help libraries to anticipate the effect of adding a demand-driven component to their ebook acquisition strategy.