The publishing industry deals with a lot of big numbers, from bestselling books selling millions of copies to combined industry book sales of tens of billions of dollars. I maintain a reference page about book sales which I update on a regular basis from census bureau numbers and the annual reports of the chains and Amazon. But the questions I get tend to make me suspect many authors and journalists have trouble with the magnitude of the numbers, which makes it difficult for them to draw the sorts of conclusions for which I'd hoped I was providing that raw data. Coming from an engineering background, I can hardly read a newspaper without finding that a journalist or proofreader has made an error of three orders of magnitude (1000 times), and nobody seems to care because it's just more big numbers.
The basic unit to keep in mind when dealing with practical large numbers is 1,000. If you don't mind rounding a little, which is always fair when doing estimations, take a 1,000 as the number of times you walk your dog every year if you usually get out three times a day. It's not an unimaginably big number, and it's certainly one you can do basic math with. If you put five dollars in a jar every time you go out to walk the dog, at the end of the year, you'll have around $5,000. Of course, while you're out with the dog, somebody may steal it, but that's a separate issue. A thousand thousands is a million. If there are a thousand people in your town walking their dogs three times on most days, by the end of the year, that's a million dog walks. If everybody puts seven dollars in a jar before they go out to walk the dog, by the end of the year that's seven million dollars ($7,000,000). A thousand millions is a billion. If there are a thousand towns in America where a thousand people walk their dogs just under three times a day, that's a billion dog walks a year.
The basic publishing industry numbers are of the same magnitudes. Ignoring the mass merchandisers (supermarkets and discount stores) who sell small selections of bestsellers, romance and life style books, there are a limited number of outlets for trade books in the US. There's Barnes&Nobles at around $4,000,000,000 (four billion), Borders at around $3,000,000,000 (three billion), and Amazon at around $3,000,000,000 (three billion). The college bookstores do another $4 B ("B" is a friendly short-hand for billion), and the independents all combined probably do the same. The leftover couple billion from the Census Bureau number of around $16 B (ex-Amazon) is taken up by specialty bookstores, primarily religion and professional books.
So, where do these big numbers get us in terms of our basic dog walking unit? If we add the Amazon number back into the Census number, and drop the college bookstores and specialty stores that most authors books aren't stocked in, any more than they are stocked in mass merchandisers or specialty shops, we're left with a mainstream bookseller number of around $14 B per year. Don't mind the rounding, it only gets worse. If we took the average price per book sold as a convenient $14, that would give us one billion books a year sold through mainstream bookstores. Or, we can say that if one thousand Americans in one thousand towns buy a book for every time they walk their dog in a year, they've purchased the total number of books sold by mainstream book retailers.
Whether you think that one billion books number is large or small depends on primarily on whether or not you own a dog and enjoy walking it. In this case, I did a lot of rounding and took an arbitrary price for the average book, but I hope, if you need to do some calculations using my raw data for book sales you won't let all those extra zeros get you frazzled and concentrate on the simple ratios. Just because numbers are big doesn't mean you can't get meaningful information out of them. Another way of looking at the one billion books is by estimating that there are less than 5,000 bookstores involved in selling them (Amazon US is one big store). That comes to an apparently meaningless average of 200,000 books sold per bookstore. Meaningless, because the stores vary greatly in size, but meaningful because it coincides with the roughly two full inventory turns a year one hears is the norm for a superstore with 100,000 books.
So I hope I've conclusively proven that while dog years are much shorter than people years, it depends greatly on the breed and how often they get walked. Besides, it's not how many years you live, it's how many books you chew.