Nathan Bransford, Author

Wednesday, May 16, 2007

How Could the Publishing Industry Become More Scentific?

As you know from Monday's post, the New York Times recently published an article that stated that the publishing industry does not devote as many resources to market research as other industries, and implicitly connected this to the fact that there is an incredible amount of uncertainty and guesswork in publishing. So this got me to thinking (watch out). Let's say the publishing industry DID do extensive market research and reader surveys and the like. What do you think they should research?

Please forgive me for a moment as I speak in the second person.

You're the CEO of a large publishing company (congrats, btw). Your company spends a great deal of money on advances for a whole bunch of books, but at the end of the day no one really knows what makes a book a bestseller. You have a young whippersnapper researching/analysis team and a substantial budget. Your goal is to remove as much uncertainty and guesswork from the publishing process as possible and, of course, increase profits.

What do you tell your young whippersnapper researching/analysis team to do?


The Anti-Wife said...

Okay, speaking as an ex director of marketing research and sales analysis, you need to know several things.

First, what are people buying and reading. This would require records of sales from many outlets - publishers, bookstores, resellers - just about anyone who sells books. You might also want to check libraries to see what books are being checked out (if they keep records).

Then you would want to break the sales down into genres and sub-genres and sub-sub-genres if possible.

Second, you need to do some research to determine what people might be interested in reading in the future, why they pick up specific books, what influences their purchase decisions, etc.

Using statistical analysis, you could determine many of the variables that factor into the sales.

However, it's important to remember that while the data may give you information on past and current trends and help in forecasting sales for the future, nothing will be 100% accurate because you are dealing with human beings with various levels of intelligence and experience.

Nothing will be perfect, but if the data is available and the research is properly structured, it is possible to make fairly accurate predictions for the future.

Nathan Bransford said...

the anti-wife-

Those are great points. The new Nielson Bookscan service provides a lot of the data that you are speaking of, and they do break things down by small genres. I do think this could make things more efficient in terms of discovering genres that are selling more than expected, but as you say, it's a very reactive medium -- you're only seeing what is already out there and what worked in the past.

The numbers, though, can contribute to a mindset where people believe an author or genre will sell about as many copies as the last book, when, as anyone in publishing knows, that's almost never true. Dog memoirs could not sell, not sell, not sell, and then all of a sudden MARLEY & ME comes along and sells a bazillion copies.

I would love to know how that unexpected success could be predicted, because you wouldn't find it in previous sales numbers.

Merry Jelinek said...

I've got to say, I don't think it can be predicted. You can't necessarily go by genre or topic, because you're not adding the voice and writing into the equation.

Okay, before Harry Potter, who would have looked at the past statistics and said, "yeah, that's what we need is a ya series about an orphaned boy wizard battling evil..."? Possibly, they might have looked at ya fantasy and noted a market existed, but they couldn't have known the scope of sales simply by the topic...

There are so many other variables - like, oh, I don't know, the writing... and good novels will be a lot deeper than their blurbs lead you to believe. How do you accurately predict the type of writing, art, film, that will touch a nerve and draw the reader in...

That's the thing here - ideas are overdone, plots have been retooled infinitum... the most ingeneous, marketable idea a writer can come up with has already hit the shelves many times over in multiple forms... It's the voice, the writing, the timing, and, for my money, the characters that drive the fiction. That's what the reader will fall in love with... that's what makes them buy extra copies for their friends and recommend it all around.

That's the beautiful thing about fiction - the minute it hits the shelves the words no longer belong to the author... they belong to the reader - he's the important element... That reader, if he owns those words, wants to pass it on to everyone he knows - if for no other reason than to bask in the glory of a fabulous find.

There's my two cents...

The Anti-Wife said...

It would require extensive marketing research - focus groups, questionnaires, surveys, etc. You would have to determine WHY people buy books.

The problem with this versus testing a product that sits on a store shelf - like cereal or lotion - is the subjectivity involved in the decision to purchase a book. I'm guessing this decision is often personal and spontaneous. Those would be difficult things to quantify, but with properly structured research you could make some decent predictions.

Scott said...

This kind of scares me.

I think the marketing department would do tons of research and determine two things:

* Established best-selling authors sell books.
* A lot of people are reading Harry Potter books.

Therefore, we need to convince Tom Clancy to write a book about a prep school for young wizards and witches. And, for heaven's sake, don't spend any money on an unproven writer; however, if you have to bring up a new writer for whatever idealistic, money-losing reasons you might ave, at least make sure he or she is writing about a prep school for young wizards and witches.

I think market research can determine what's selling and what's likely to sell based on what's hot right now, but it can't predict The Next Big Thing, and, in fact, could prevent The Next Big Thing from happening.

A really good marketing team could create a perceived need for a certain kind of book and push the company to fill the hole they invented in the market (most likely with proven authors), but I think that's about as close as it can come to predicting what will sell in the future.

Either way, a new breakout trend or author has even less of a chance than now.

At least that's my own cynical view based on experience in a different industry.


Though it would be subjective, an agent or editor could discern a difference between Marley & Me and the not sell, not sell, not sell dog memoirs. The quality of the writing or the nuance within the concept could give a clue.

Christopher M. Park said...

Here's an idea for a potentially (comparably) simple way to gather this data: why not ask the big bookstores? They probably already have a lot of this data.

Back in high school I work in an Ace Hardware for a year, and we had about 5,000 products on the shelves (that sounds like a lot, but a large part of that was simply small plumbing parts, nails and screws of various sorts, etc). The managers of that particular store would order their products from the central Ace Hardware catalogue, which in turn had about 25,000 products available. It was mostly up to the managers to decide which of those products they should carry for their specific area, and they did this by analyzing detailed POS data (that's "Point of Sale," not the other thing) and by making intuitive guesses. When an intriguing new product came along in the catalogue, the managers or owners might order a couple and set them in a prominent place to see if they sold. If they did, they'd order more, and over a period of a few weeks or months they'd figure out what the ideal stock number was (maybe zero).

Presumably you see where I’m going with this. I’m pretty sure that bookstores follow much this same pattern--so far as I know, they have a lot more detailed, concrete data than, say, the publishers or agents do. I don’t believe for a second that the big bookstores are any less scientific in their business model than any other retail chain in another industry.

So, that’s where I would go to get the data if I didn’t have that big staff and budget that you mentioned (and it sounds like that is closer to reality). The bookstores might not want to share this sort of information, and/or their privacy policies might make it difficult to do so, but in the end it seems like it would be beneficial to them all if publishers were printing more of the kinds of books that they wanted, and less of that which they don’t want to buy anyway. I’m guessing that the big bookstores have all this data broken down by region, genre, author, you name it—and if not, that’s where I think the first investment should be made, in better centralized POS tracking and inventory management systems at the major booksellers.


Liz said...

Not sure if this relates or not, but I'm going to pass it along as it was something I'd never seen before.

This past Sunday's newspaper came wrapped in the regular plastic bag, but attached to the outside was a Harlequin romance mini book.

The 'book' contained the first two chapters of four different novels. At the end of each section a discount coupon for the complete book was inserted.

Now, granted, I do live in a town that could pass for Stepford, but I've never seen a publisher do such direct marketing to suburban housewives. I've seen laundry soap and shampoo promoted this way...but books?

John Pettitt said...

This is the same problem technology companies face. You really have two paths you can choose from. Path one is to look at what is selling and do more of that. It’s not a bad strategy and it’s one that lends itself to lots of hard data. The problem with this strategy (at least in tech) is it almost never leads to a break out product. The problem is that customers don’t know what they want. For example no amount of focus groups or statistics could have predicted the iPod. This leads to path two: make leaps of faith based on experience and intuition. I suspect that focus groups and market research would not have supported a conspiracy theory book about the Catholic Church. However somebody made a leap of faith and the rest is history.

In the end almost every product strategy comes down to one of the paths above.

Laurel Amberdine said...

I would fire the researching/analysis team, except for an intern or two. The experience of reading is an emotion which can't be quantified without actually reading stuff.

I would tell the surviving interns to find me a bunch of avid readers of each genre -- search blogs or livejournal or go to conventions. Heck, take out an ad. I'm sure it would be trivially easy to find plenty of widely-read intelligent, passionate people to read and comment on books. (I suspect this is what agents and editors functionally are.)

Anyway, neverminding that, and because I have this big budget to spend, I'd interview and filter my readers to find a bunch of people who read a lot, and tend to like things that sell very well.

Keep feeding them books. Give them blogs to blab and get comments from other readers. Find out what they like, what they're tired of, what they want to see more of.

One detail, though... make sure they're not writers. Writers are too picky and generally hate bestsellers. ;)

Len said...

As the CEO of Leviathan Books, I would tell the marketing staff, in the British parlance, to piss off. As Hollywood has shown us, you can survey and do focus groups and all that stuff, but all you end up with is blander fair that succeeds no more often than the spawn of the "gut feeling" method that prevailed before.

When it comes to which product cleans better or what packaging is more attractive, you can get usable information. But as to what makes someone laugh or cry or think, that's a whole different can of worms.

alternatefish said...

hey, does anyone know how the music industry deals with this problem? I feel like music and publishing kind of parallel in the way they need to deal with the uncertainty of human taste in order to sell.

Stephen Parrish said...

I'd hire Christopher M. Park.

Stephen Parrish said...

By the way, Nathan, the publishing industry could become more scientific by spelling the word "scentific" correctly.

Sorry, couldn't resist. You rule.

Nathan Bransford said...


Haha, true enough. I'll leave in the error so as not to mess up people's Bloglines.

Happy Days said...

Oh, my I hadn't even noticed.

Well in that case, the answer is obvious: Use Scratch and Sniff test marketing methods.

(I know. It's a groaner.)

Anonymous said...

It can't and it shouldn't. It should get *less* scientific and only resort to marketers when trying to sell a book that's already been bought by the publisher, rather than resorting to marketers to try to decipher what books ought to be bought in the first place.

Peter R said...

Yes, the publishing industry could become a lot smarter and scientific, but not by trying to predict future trends. As everyone correctly points out choosing a book is far too subjective: it is an emotional process influenced by trend, history, current events, recommendation, who won X-factor, etc. All impossible to predict two years into the future. However, where scientific and statistical techniques could be better used, and Christopher Parks has already given a crude example, is in identifying emerging trends and predicting future sales of current publications. DirectLine balance the risk exposure of their insurance customer-base by the hour: if they have too many young drivers at 10:00 am, they drop the price for older drivers and raise the price for younger drivers until the balance is restored. If the key emotional factors are know for why people buy certain genres and types of books, then those key indicators can be tracked for an emerging trend. Customer loyalty cards work really well in this respect – a great source of data. When a rising trend is spotted the marketing people could weigh in with additional promotions, and the printing presses start rolling. It will never be an exact science, but as a business model it has to be better then printing thousands of copies and investing loads in marketing in the hope a book will sell, then pushing them onto the cut price shelf to move them on when they don’t sell. I reckon that some version of print-on-demand has a great future in the large publishing houses. As an author this all probably sucks though.

Don said...

To address alternatefish, the music industry became more scientific and their product suffered and now people buy less music because it's all scientifically marketed.

whitemouse said...

I suspect it would be a waste of money, because I think you'd find that people like to read (*gasp!*) really enjoyable books.

("Enjoyable" being distinct from "well-written".)

Having discovered that, what would the industry then say to writers? "Write a really enjoyable book."

Which, um, most of us were already taking a stab at doing.

Simon Haynes said...

Problem is, the book-buying public knows what they love when they read it. Before that it's guesswork and a lot of 'if you liked A you might like B' - even though no two books are alike.

Lisa said...

I agree with John Pettitt's comments. In my industry, market research tends to focus on historical trends and history rarely, if ever repeats itself. Marketers also tend to "predict" trends that aren't being adopted nearly as fast as they'd like them to be because they've invested in a prediction. Marketers are typically averse to asking customers what they'd like to see because it's expensive and because they'd rather tell them what they should want. As JP noted, the customers have some idea, but they can't predict what they'll love if it doesn't exist yet. I suspect a lot of readers are like I am and could explain what they'd like to see more of. I'm in a minority, but if there was more literary fiction with strong female characters (no more victimized, abused, single mothers who need to be rescued or have to overcome all their issues) I'd read it all. One day there will be a John Updike or Wallace Stegner equal with flawed, mature female characters.

alternatefish said...

ah ha. I see. that explains a lot. so there'd maybe be a lesson there, yes?

Related Posts with Thumbnails