Hey Felix, I had no idea you were such a gallant defender of the Times’ bestseller lists. The surprises never stop.

To take on your comments.

1. You ask how many medical/law/textbooks/etc would sell fast enough to make it onto a bestseller list. I dunno, and you don’t know either. As far as I’m aware, no one knows. Which is kind of my point – that the picture that mainstream books coverage gives of the book industry is often more than a little misleading.

2. Great to hear that you understand the way bestseller lists work, and what the implications of that are. I’ve found that many people haven’t had a chance to give it much thought, and that’s why I’ve put up this posting. I’ve found that many people have no idea, for example, that it’s perfectly possible that a book might sell a half-a-million copies yet never show up on a bestseller list.

3. I’ve got nothing against anyone looking at any bestseller list they want to look at, but I do like supplying a little info about them when and where I’ve got a little trustworthy info to provide. There are a bunch of interesting lists people can check out – in Publishers Weekly, the WSJ, B&N’s own list, the Voice’s indie-bookstore list … It doesn’t hurt to know what goes into a list, though. The Amazon list, for instance – fascinating in its own right, but there are a few things it can be useful to keep in mind. One is that Amazon, despite its huge presence in the public mind, is just one retailer. I don’t know any longer how big a piece of trade-book sales it accounts for, but as of a couple of years ago it was less than 5%. Also worth keeping in mind is who shops at Amazon – people with computers, and who are comfortable buying online. An interesting demographic but a peculiar one, and not all that representative of the country at large, at least so far. If you take the Amazon list as gospel, you’d be missing out on something big -- the fastest-growing part of books retailing, which is bigbox bookselling. Places like Costco and Sams Club, which carry a very small number of titles, now account for something like 25% of trade book sales. Amazon’s rankings won’t show any of the effects of that fact.

4. Glad we agree that methodology will always be imperfect in a case like this, although I do think if/when BookScan is in place it might well put an end to the rivalries between bestseller lists, as MusicScan put an end to lots of speculation in the musicbiz. It’s fun to poke and prod a bit to draw out the imperfections, though, isn’t it? Why not have a good time figuring out in what ways a given methodology is imperfect? Amuses me in this case, anyway. And, yeah, I’d say (without having given it a lot of thought) that bestseller lists are probably more misleading than television ratings are. For one thing, I suspect tha TV ratings don’t ignore 2/3 of the TV market. I suspect bestseller lists are more misleading than movie-gross lists too. Why? Because while many books have long, long lives – a fact that a bestseller list usually can’t portray -- movies usually do the vast majority of their business in the first year. Your point that anyone who wants a really complete picture of a given business has got to consult multiple sources couldn't be more true. The more context, the better.

5. Well, since I specifically say that I’m not accusing the Times of bad faith, I don’t understand why you seem to be arguing as though that’s what I’m doing. I do see two issues in what you’re raising, though. One is the rationale behind how the Times’ lists are broken up, and the other is whether or not the Times is doing anything to make it appear as though classier books are selling better than they really are.

As for the first, I’ve got no idea how they arrived at their current lineup of lists. I picture tons of conference-room meetings, but have no info here and so will shut up.

A related topic: You (and many other people) may enjoy it that the Times has gone to the trouble of segregating out softcover-advice books from, say, hardcover-fiction books. I find it bizarre. It may seem natural to us these days that bestsellers should be sliced and diced into a certain number of now-familiar categories (hard/soft, fiction/nonfiction) – but I ask: why these categories? I note, by the way, that one of the Times’ categories is Children’s Picture Books. Odd choice: Why break that category out? Was there a Mommy in that conference room who had to be placated? But there are literally dozens and dozens of ways of slicing and dicing the world of trade book sales – mysteries, computer-advice, romance, pop-science, cookbooks, you name it. Large-format/small-format. Heavily-illustrated/ largely-text. Below-$15/above-$15. All of which seem as legit to me as the categories the Times uses.

As for the second: Is anything going on at the Times to make it appear that classier books are selling better than they are? Sure, though this used to be much more flagrantly the case than it is now. The Times admitted as much when they reformed their lists five-ish years ago. The Times also admitted at that point that part of the reason for hiving off children’s books was to get the “Harry Potter” books off the hardcover-fiction list. Interpret that move as you will. I’ll pass on the chance to reveal my sources, though. Feel free to disbelieve me if you want to.

6. A. Aren't we saying the same thing here? That you, er, one sees a rather different picture depending on whether one sees a book like the Candace Bushnell in the context of trade-book-sales generally, or lifted mostly out of that context. I prefer the former; fine with me if you prefer the latter. But could you have predicted that the Bushnell, #5 on the NYT hardcover fiction bestseller list, would show up at #47 measured overall? I couldn’t have either.

B. The hand you wave at the “Harry Potter” books, some of the most popular books ever published, seems to me a little casual. I don’t agree that dismissing their popularity as "boring" makes any sense, at least not if what you’re in search of is objective fact. What does boring or not-boring have to do with providing a clear picture of which books are selling how fast?

C. It’s not exactly a big surprise that the hardcover-fiction bestsellers should appear in the same relative order on the two lists, is it? Who would expect otherwise? My point is that the USA Today list offers a lot more context and information.

Because you got my goat a little, I went to the trouble of printing out and comparing the lists from both papers.

Here’s what you learn from the Times: that “Da Vinci Code” is #1 on the hardcover fiction bestseller list; that “Johnny Angel” is #2; that “The Lake House” is #3; that “White Death” is #4; that “The Lovely Bones” is #5. That’s it. That’s all you learn. How do their rates of sales compare to Atkins, to Hillary, to “The Power of Now”? No way of telling.

Here’s what you learn from the USA Today list: that “Da Vinci Code,” while indeed the bestselling hardcover-fiction book in the country, is only the sixth fastest-selling book overall (and comes one after the Oprah-selected “East of Eden” in paper); that “Johnny Angel,” while #2 on the Times list, is only the 22nd fastest-selling book overall (and is selling slower than “Kate Remembered,” “Benjamin Franklin” and a Nicholas Sparks paperback); that “The Lake House” is #25 overall (and is selling slower than a Julia Quinn romance paperback – who’s Julia Quinn?) …

“The Lovely Bones”? While #5 on the Times’ bestselling hardcover-fiction list, it’s only the 43rd fastest-selling book in the country – and comes after not just a “Star Wars” novel and a bunch of romances and diet books, but even after the paperback reprint of “Life of Pi” (#35). Which means that people who live in the world as the Times portrays it are under the impression that "The Lovely Bones" is hot hot hot --quite the phenom, and on everybody's lips. People who follow the USA Today list know that, while the books has done awfully well, god knows, at the moment it's not selling quite as fast as "The No. 1 Ladies Detective Agency" or "The Atkins Journal Package."

I dunno. You may not find any of that interesting; I do. But interesting or not, it’s certainly a lot of extra information, and a lot of extra context as well. And it's why I recommend looking at the USA Today list.

Imagine yourself standing at the door of a bookstore, and you’re checking people’s shopping bags as they walk out, and you're counting copies of the books they purchased. USA Today’s list will give you a much closer approximation of what you’ll see than the Times’ lists will.

But you’re absolutely right – why not check both lists? And Amazon’s too?

Posted by: Michael Blowhard on July 20, 2003 12:37 AM

Felix:

I think the Neilson people would have a heart attack at having the accuracy of their numbers compared to the accuracy of the NY Times bestseller list. Whatever you want to accuse them of, they certainly know their statistics, their sampling, etc. While there remain issues with their numbers, they are issues that are in reality very difficult to deal with. Whereas, from everything I've heard, the NY Times list would be instantly rejected as scientific evidence of anything, being largely assembled on a wish and star.

Posted by: Friedrich von Blowhard on July 20, 2003 1:58 AM

I feel like I'm being excluded from some kind of cult here! Michael says he will "pass on the chance to reveal [his] sources," while Friedrich simply cites "from everything [he's] heard" as evidence for the NYT list "being largely assembled on a wish and star". Is the Times, even today, so powerful that people are reluctant to criticise it in public, at least beyond the confines of a pseudonymous website?

For all of the heat this discussion has generated, I think we're more or less in agreement on substantive issues. Michael, you obviously care more about the Harry Potter decision than I do, but I thought it was a bad decision too. It's just that I can understand it: when you have limited space to publish the list, and the top half is all Harry all the time, you're not giving people information on how things are changing -- which is what a weekly bestseller list is surely all about.

I really do understand your desire for an aggregated all-time bestseller list. I'm sure we'd see a lot more Alice in Wonderland and a lot less South Beach Diet. But that's a once-a-year list at best, not a once-a-week list. What's more, I have a feeling that it would be a bit like the all-time movie grosses list: biased towards recently-published blockbusters. Why? The very bigbox phenomenon you cite: while overall book sales are growing relatively slowly, that hides huge growth in Wal-Mart and slowdowns in your friendly local crime-fiction store. Increasingly, the publishing industry seems to be concentrating on home-runs rather than slow-but-steady sellers: note that your Alexander book, at sixty-some dollars a pop, is published by a university press and not one of the profit-driven giants.

You do indeed say you're not accusing the Times of bad faith -- but in the very next paragraph go right ahead and do just that, in your vaguely-sourced "it's generally thought that the reason the Times massages its lists in the way it does is to give books that the Times deems to be respectable a better chance of appearing to be bestsellers." That's an almost textbook definition of bad faith, isn't it? Publishing something purporting to be an objective list of best-selling books, but which in fact is massaged towards one's own preferences?

You write: "one sees a rather different picture depending on whether one sees a book like the Candace Bushnell in the context of trade-book-sales generally, or lifted mostly out of that context. I prefer the former; fine with me if you prefer the latter. But could you have predicted that the Bushnell, #5 on the NYT hardcover fiction bestseller list, would show up at #47 measured overall?"

Exactly at #47? No, of course not. But if a book was at #5 among hardcover fiction, would I imagine it to be much higher than 47 overall? No. Look at how much space hardcover fiction takes up in the average Borders: it's a tiny proportion of book sales, obviously, and the only reason that the top sellers do as well as they do is because of all the successful hype around them.

What I don't understand is why you want to lump hardcover fiction sales in with softcover economics textbooks and whatnot: we're completely in the realm of apples and oranges here. If you go to Amazon, does their bestseller list combine books, DVDs, cellphones and office products? No, all those lists are broken out, because it really isn't useful to know if Candace Bushnell is selling more than the Office Depot Moving Box, 3-In-1, 12in x 9in x 6in. The Times, understandably, I think, wants to compare like with like as much as possible. What useful information would we get if a bunch of economics textbooks suddenly appeared at the top of the list every September, while the mathematics textbooks didn't? Basically, nothing more than the fact that economics is a more popular college course than mathematics is. And for that we would have to search even harder than usual for the kind of books which have broad appeal and are popular because people really want to read them.

Posted by: Felix on July 20, 2003 10:52 AM

Felix -- Let me pinch something off before it goes too far. This posting isn't an anti-NYT-bestseller-list posting. The Times' bestseller list is what it is, and if people find it useful or entertaining, bully for them. Couldn't matter less to me.

Here's what the posting is. Two parts: A) A half-a-dozenish facts about the nature of bestseller lists, followed by B) A humble suggestion that people take a look at USA Today's bestseller list if they want to get a more realistic and representative picture of what book sales are really like in the country.

If you're saying that my half-a-dozen-ish facts about the nature of bestseller lists don't come as news to you, well, excellent. Many people, though, don't hang in arty media circles to anything like the extent you and I do. I'm hoping they get a kick out of learning a few things about bestseller lists.

If you're saying that you prefer looking at the Times' bestseller lists to looking at USA Today's, well, who's trying to get in your way?

My point at the end of the posting is that USA Today gets you several steps closer to the raw book sales data. It clearly does -- the Times, as well as most other places, puts several layers of judgment calls between you and the data. USA Today keeps the judgment calls to a minimum. You may not like that; you may enjoy having a few judgment calls put between you and the data. But that's a personal preference.

I don't quite understand why this sentence seems to hypnotize you so: "it's generally thought that the reason the Times massages its lists in the way it does is to give books that the Times deems to be respectable a better chance of appearing to be bestsellers." Although I can't figure out why you're so fixated on it, here's some substantiation: I lunched with publishing people a couple of times a week for 15 years. Bitching and speculating about what the Times might be up to with their bestseller lists is a standard topic of conversation among publishing people, and the interpretation I pass along here is what most people, in my experience, seem to have settled on. Perhaps they're all wrong -- who knows? They don't know, you don't know, I don't know, and the Times keeps aspects of their formula a secret, which is their right. It's certainly true that the Times' bestseller lists used to be assembled by an old lady whose research consisted of making phone calls to friends at a select number of bookstores (an exaggeration, but only a slight one). And the Times does deserve credit for reforming its lists about five years ago. They're much more professional and open than they used to be.

But like I say, this post isn't a slam at the Times' bestseller lists. As long as you know what they are, and as long as you have a pretty good sense of the context their results and presentation exclude, they can be useful and interesting. I'll note that the Times doesn't dwell overmuch on the nature of their own bestseller list, and they don't provide much in the way of context either. Which is where reading 2Blowhards can come in handy.

Posted by: Michael Blowhard on July 20, 2003 2:29 PM

You're right: As I say, there's little substantive difference between us. You make a good point that the USA Today list has more information than the NY Times list, and generally that's a good, not a bad, thing.

And even more generally, maybe you're right about levels of sophistication. I remember when I was 11 or 12 I would look through my father's paperbacks and see many of them with "New York Times #1 Bestseller!" plastered all over the front cover; it confused me, then, how all these different books could all be number one.

Anyway, I'm with you on BookScan. Here's hoping they make all the data available on the internet for free. Pigs might fly, eh?

Posted by: Felix on July 20, 2003 3:27 PM

Hey, they might!

Posted by: Michael Blowhard on July 20, 2003 3:34 PM

One more thought. I recall reading, some years ago, that the NYT list did not include sales at religious bookstores. Some religous books, like those in the "left behind" series have had quite high sales from other reports I have seen, but never make the NYT list.

Posted by: Jim Miller on July 20, 2003 8:41 PM

I don't know about bookstores, but I have actually seen record store employees (back when there were record stores) manipulate numbers for the bestseller list. When I worked at a record store near the University of Chicago 25-30 years ago, we would get a call once a week (I think) from Billboard (I think). More than once the discussion would go something like this: A. "Don't tell them Blue Oyster Cult! That sucks!" B. "But we sold five copies last week." A. "Tell them [sorry, I can't remember a specific example]. That's cool, and we sold at least two or three copies." And we would tell them the cool one that sold less instead of the crappy one that sold more.

Posted by: Dr. Weevil on July 20, 2003 10:08 PM

I happen to work at an independent bookstore where I am in charge of reporting our sales numbers to the New York Times. It goes like this: They send over lists of titles broken down into the catagories that they use, hardcover fiction, paperback advice, etc. Then I go through our computerized inventory system, track down how many we sold in the past week for a particular book, and write that number on a line next to the title. I then get the computer to print out titles that have sold well in the past week and sift through them looking for things to add to the blank lines below their lists. Often times we sell great quantities of books that do not appear on the lists they send me. Then, finally, I fax them back the list with my quantities. All in all, this process is time consuming, prone to human error or deceit, and not all that different from the old lady who called her friends at various book stores.

Posted by: Max on July 21, 2003 12:25 AM

Perhaps this post was stimulated by the recent splendid piece by Tim Adams in the Guardian, where he reprises the Gore Vidal/Anthony Lane acts of bravery by reading all the books on the current bestseller list. His conclusions, as you might expect, are amusing. Interestingly, he notes that bestseller lists are a recent development in the UK.

I know what these lists mean to marketers and retailers but I do not think they mean much, if anything, to readers. What reliable, useful information does one know about a book when one knows its sales ranking? Are there even sociological and cultural conclusions that we can draw from sales trends?

Joseph Epstein was thrilled that Snobbery: The American Version was a bestseller. Now that he is on a book tour for Fabulous Small Jews his publisher is putting him up at the Fours Seasons. I guess bestseller lists are good for something.

Posted by: Robert Birnbaum on July 21, 2003 8:03 AM

Jim -- You're absolutely right. Now that you mention it, it comes back to me that the standard bestseller lists don't give enough weight to religious books -- or to certain genres that are sold in specialty stores. I did a little Googling to see if I could verify this and turned up a good Clive Thompson article from the WashPost (reprinted for some reason here). Bookscan, books that are shut out from the bestseller lists -- plus some other details, stories and quote. The equivalent of Bookscan is operational in Australia, for instance -- and it's had quite an impact. A good line from Thompson about whether the lists are massaged: "Not that major newspaper editors particularly deny this. As most admit, their goal is not to replicate reality but to provide a snapshot of the culturati."

Dr. Weevil, Max -- Garbage in, garbage out, eh? Great info, thanks.

Robert -- I hadn't run across the Tim Adams piece, no. I've now read and enjoyed it -- thanks for pointing it out. It's here.

Posted by: Michael Blowhard on July 21, 2003 10:50 AM

You might be interested in Barnes and Noble's year-to-date bestseller list. It doesn't include textbooks and it's only one vendor, but it does cover a much longer period of time. Of course, statistically it would be much better if BN used the previous 12-months' worth of data, rather than year-to-date figures. YTD means the list starts out just like a weekly bestseller list, then slowly morphs over the year into an annual list. But, especially at the end of the year, the list is much better at capturing overall sales.

Incidentally, the BN lists do not just cover the computer-comfy shoppers (as on amazon)---book sales in the actual stores are included in the data.

Notes on earlier points:
Friedrich is quite right that the TV viewer numbers are much more accurate and reliable than bs lists. There are two quick reasons for this. First, the TV data estimate actual viewers---bs lists don't tell you how many books were sold. Second, in the "that's where the money is" category, TV "eyeball sales" are directly related to the major revenue stream of TV: advertising dollars. You can bet that if books were free to readers, but Ford and Clorox paid for color ads in the middle of the book, there would be not just accurate numbers on distribution, but also how many people actual read the book.

This is an issue that is largely ignored by looking at bs lists. What is the relationship of book sales to book reading?

Posted by: Louis Hicks on July 22, 2003 1:08 AM

The English author Claude Cockburn wrote a book about bestselling novels many years ago (I think it was called just BESTSELLER from memory) and he stated that he only covered books up to the 1940s because after that publishers and publicists became too involved with bestseller lists. Doesn't sound like much has changed.

Posted by: Mike on July 22, 2003 2:14 AM

Given that the "accurate" list probably has little of interest in it for the average NYT reader, I suspect that publishing the "accurate" list would shortly result in the removal of the feature altogether.

Harry Potter + Atkins + Left Behind = Boring.

There's a reason that TV ratings aren't a common feature in most publications. Who cares?

I don't think that there are too many people who take the NYT list literally. It's sort of a combination of selling a lot of copies with some literary worthiness. Putting "NYT Bestseller" is a lot more meaningful to this reader in many ways than "Sold a bajillion copies".

Of course, if I was an author, I'd make awfully certain my book was the right kind of book before I put a "NYT Bestseller" clause in there.

Posted by: Tom West on July 22, 2003 11:58 AM

"Was there a Mommy in that conference room who had to be placated?"

How condescending. And what about all the daddies in the conference room?

Posted by: R. Forster on July 23, 2003 12:09 AM

Louis -- Thanks for pointing out the B&N list, which is very interesting, as well as quite a contrast to B&N's "hourly bestseller list." The year-to-date list is here.

Mike -- I think I'd probably be less prone than Claude Cockburn to accuse anyone of mendacity. The bestseller lists are products, and as long as people understand the nature of the product, I don't see any trouble with it. And, like I say, as far as I've been able to tell, the USA Today list is probably as accurate and full of information as such a list can be, at least until Bookscan has the country covered.

Tom, that's a fascinating p-o-v, thanks. I guess I'm a little perplexed that the misleadingness or massaged quality of the Times' lists wouldn't bug you, but pleased to hear you find them useful. You really wouldn't rather have accurate data straightforwardly presented?

R. -- I've sat in on hundreds of media-world meetings, and the person who brings up such a let's-remember-the-kids idea is invariably a Mommy. Me, I'm usually glad she's doing it. What's condescending about any of this?

Posted by: Michael Blowhard on July 23, 2003 1:26 AM

While bestseller lists attempt to tell us what books people buy, they don't tell us what they read. A few years ago, Newsweek,of all places, did a small Top Ten of what people were actually reading. A comparison with Bestseller lists revealed no contiguity whatever. Seems, according to the writer, that people felt it important to buy Stephen Hawkings' "A Brief History of Time" but that most buyers set it aside "to read later" and never broke the cover. Add to this the snobbish purchasers who placed the book on the living room coffee table to impress guests, and you end up with little idea of what is actually moving and shaking people. If you would like to believe that we are ascending to ever and ever higher planes, then standing at a bookstore cash register for half an hour (even in a tony bookshop) and watching what people actually pay cash for, is a sobering experience that bears scant resemblance to the items on bestseller lists.

Posted by: Steve on July 23, 2003 2:10 AM

I workede at B&N for two and a half years and can say truthfully that their bestseller list is pretty much rigged. They make deals with publishers, based on preorders and numbers of copies printed. If Publishing house A is hyping a new author they make a deal with the distributor to give them a discount on bulk orders. That's why when you walk into a B&N store, there are five hundred copies of the new Dr. Phil book laying around and it debuts on the best seller list. It's marketting. They aren't books that are selling well, but books that the publishers and B&N want to sell well because they invested in them and have 50,000 copies sitting in a dozen warehouses.

Posted by: Keith on July 23, 2003 11:44 AM






Post a comment

Name:


Email Address:


URL:


Comments:


Remember your info?