[Elfwreck]

Originally Posted by Worldwalker

@Elfwreck, that still leaves the problem of sorting out a difference in sales based on illicit downloads versus a difference in sales based on, say, reviews for the book.

Definitely.

It's not a perfect system, & it'd need a lot of tweaking. There's a lot of uncounted variables.

It'd be a *start*, a way to begin collecting data about how piracy actually affect sales, instead of blithely claiming that every download is costing the company a new hardcover's price.

That's the big problem with any attempt at teasing out one variable out of the many that affect book, music, movie, or any other sales: it's impossible to have a control group.

The control groups are "books with no legit ebook edition that we can't find pirate versions of" and "books with legit ebook editions that we can't find pirate versions of." (I own several in-print books in category 1; I'm less sure if the others exist, but certainly "books with legit ebooks whose pirate versions are really hard to find" is possible.)

I freely grant that it's a very loose, rough way to gather data. It won't get nice chartable numbers; at best, it gets general-overview data indicating trends, not hard fact points. It's *possible* that more specific tests could be established if publishers did some general data-searching, but it's also possible that there are just too many variables.

The publishers who've admitted to doing the general data-checking have come back with reactions like Baen's: "Pirates? You mean, those people who are telling other people that our books are worth reading?"

There are just too many factors, such as reviews, marketing, word of mouth, etc., that are individual to any given book.

Can't get (useful) data for individual books, but tracking data across a couple dozen or a couple hundred of them would show whether there are trends. If the data showed "hardcover sales went up after ebook release, even when ebook sales were pathetic," that indicates that piracy might increase hardcover sales. The real fun is trying to track connections... do Book 3 Of Series sales increases when Book 1 is released for as a free ebook? (Baen knows the answer to that one. Macmillan is apparently convinced they lose too many Book 1 sales to bother with it.)