Customer reviews are the latest buzz in marketing:
Problems have turned up with customer reviews in all these areas and for obvious reasons. Chief among them: corruption. It is understandable that in a highly competitive market, all shots are allowed. If not allowed, at least tried. Friends will write favorable reviews. Businesses will be tempted to provide incentives (say a discount) for good reviews.
But – there’s always a but! – if the customer review system is broadly corrupted, it collapses. Nobody believes in it anymore, potential clients run away. The best of marketing tool suddenly boomerangs, and turns into the worst.
So those who run customer review systems have to be very, very careful and constantly weed out the offenders.
The Amazon Solution
Amazon is probably one of the most adept at this sort of game – mainly because it uses a double prong.
On the one hand, it has set up boxes for customer reviews (with one to 5 stars) as well as buttons for “likes” just below the book title (for those who can’t be bothered to write reviews) and buttons for “tags” (to indicate agreement with descriptive keywords that help in searching for the book – you the reader can also provide the tag you feel best describes the book).
On the other hand, Amazon keeps track of every customer’s past acquisitions. In other words, it uses social feedback and buying behavior. That’s what I mean by a double prong. What doesn’t come out right with one is corrected by the other.
Very clever, and no doubt it explains to a large extent the success of Amazon as a virtual bookstore, where much of the book discovery work is already done for you, and you are directed into virtual shelves that contain books similar to what you have bought in the past and that correspond to your tastes.
Trouble is: what happens if you want to get out of your past purchases and try something new? It recently happened to a friend of mine: in a bout of enthusiasm, she had filled her newly-bought Kindle with all sorts of 99 cents books, just to see what they were like. After a while, disappointed by most of them, she had decided she wanted to try something else. A different genre, a different price range. She knew the sort of thing she liked, but Amazon didn’t. And at that point, finding a book became very, very difficult. She should have probably started from scratch and moved into the Amazon site incognito!
Why Book Discovery is so Difficult
Actually, we are touching here the core issue of what makes book discovery so very difficult.
Few books are actually genuinely “discovered”. Trouble is: social feedback – i.e. book reviews – feeds on itself.
What drives public awareness of a book are marketing dollars – especially if movies are involved. When a movie is made out of a book, it becomes the buzz of the town. Of course, that’s what all aspiring writers dream of: that a movie be made from their book. The ultimate consecration. But there are various degrees of consecration like the Pulitzer Prize, the Man Booker Prize etc. No matter how you view it, the book is there because somebody – usually the Big Publishers – have spent a lot of marketing dollars on it.
So in terms of book discovery, we have a very simple equation: marketing dollars drive public awareness, this in turns drives what is discoverable and recommended. Therefore bestsellers occupy the top of the heap, simply because more people know these books exist. Presumably, that leaves a lot of very good books at the bottom of the heap, undiscovered.
Can this equation be broken?
How to Make Book Discovery Easy: the BookLamp Solution
There is one interesting, experimental website that is trying to do just that: BookLamp.Org.
Arising from the “Book Genome project” started in 2003 by a bunch of students, BookLamp started working a year ago, setting up a new kind of book recommendation engine. Similar to what Pandora does for music, it is based on what it calls the “story DNA” or “data points” – really the book’s thematic ingredients, covering both contents (physical characteristics, history, environment etc) and the way it’s written (pacing, motion, dialogue, description, density).
So far, it has some 20,000 books in its database, currently tracking over 600 million “data points of book DNA” in their system (and growing!). They started out with fiction but also mean to cover non-fiction, and they work only with publishers, chief among them Random House and Kensington Publishing Corp.
What’s their objective? Simple: help readers find the sort of book they like to read. As they put it in their FAQs section: “If you think about it, what we’re basically doing is championing the idea that the contents of a book – what the author actually wrote – should be the primary consideration when finding new books for readers. It shouldn’t matter that one book has a million dollar marketing budget, and the other is from a new author with no track record at all. If the content between the covers is a good match, that should be all that matters.”
Sounds great, doesn’t it? Here is a way of discovering books independently from any marketing push, just focusing on a book’s contents. In other words, it makes the value a book potentially has for a reader the key feature, responding to a given reader’s tastes. And of course, it would be any author’s dream come true.
How does BookLamp do it? By asking you to indicate your favorite title(s), then it searches for similar stuff. The system usually works, except when some “zingers” turn up, as they call it, largely a result of not having yet a sufficiently large database to find suitable matches in every case. For example, if you have outlying tastes and give them a book title that is unique in style (say cross genre etc), they may have trouble coming up with satisfactory choices for you.
What BookLamp Needs to Achieve its Objective
They admit themselves that they need an extra 100,000 books to smooth out all the crinkles and achieve their cruising speed and main objective – which is of course to sell their tool to publishers. They want to be the book Pandora. If you read their FAQs, you’ll see they repeatedly ask you to suggest publishers names, so that they can add more books to their database.
It is surprising that publishers are not flocking to them. Particularly since publishers have had a notoriously hard time figuring out the “next bestseller” and finding the way to connect directly with readers. That was, as you’ll recall, the purpose of Bookish launched in May of this year, with the backing of Hachette, Penguin and Simon & Shuster but it’s still not functional. In fact, publishers act very much like huge steamships that are having a hard time turning themselves around in the digital storm.
I don’t know why publishers are reticent and it may well have something to do with being such “huge steamships” (hopefully, not the Titanic variety!).
And I don’t know why BookLamp isn’t willing to turn to Indies: that would be an inexhaustible supply of books to test their search engine with. They needn’t accept everybody: they could call the shots and specify the types of books they want to fill the “holes” in their system. After all, the stigma of self-publication has disappeared now, with the successes of John Locke, Amanda Hocking, J.A. Konrath et al. Surely BookLamp must have taken notice?
What do you think?
What is your take on this problem of book discovery and the sort of solution provided by BookLamp?
Personally, I worry that their solution is too dependent on restrictive algorithms and that good books will still remain at the bottom of the pile. But I’d be willing to give them the benefit of the doubt, until they’ve proven, with an additional 100,000 books in their database, that it really works…
- TripAdvisor publishes its 50 millionth review (travelnews.britishairways.com)
- Social recommendation or “book DNA”? – How to find what you will read next. (nextlevelofnews.com)
- Amazon Brings Social Reading to Kindle – But Will You Use It? (readwriteweb.com)
- BookLamp Launches a Pandora For Books (mashable.com)
- On fake reviews… (shannonturlington.com)
- In a race to out-rave, 5-star web reviews go for $5 (seattletimes.nwsource.com)
2 Comments Add yours
First, thanks for stopping by my site, but I love that I’ve found yours. Very well written. I guess I’m algorithm leary. You can see it working on sites like Amazon, yet you also see it working against you on the same site in reference to your friend who bought a lot of 99 cent self-published books then didn’t like them. I guess this sort of reminds me of the “Evans-Pritchard” (Pritchett?) scale in the move “Dead Poets Society.” That scene was about ripping out the first part of a large poetry compilation (a la Oxford) because it had a mathematical chart on how to determine if poetry was good or not! I guess I remain enthusiastic, but skeptical!
Great article, very in depth 🙂