Posted on November 1, 2013 in Archive

Google might be your best bet for book recommendations


The book discovery conundrum. The large book market today means there’s too much choice. We all know about it, we all spend sleepless nights tossing and turning over how to fix this problem. How are we supposed to find a gender/age/genre/length appropriate book to bury her nose in nowadays?

In case you can’t tell, a lot of this discussion has stemmed from the WH Smith controversy over the last few weeks. That situation proves that book curation has a long way to go before it can claim that it is without error.

According to Andrew Rhomberg, there are five different kinds of book discovery:
1. Serendipitous
2. Social
3. Distributed
4. Data-driven
5. Incentivised

The first in the list is fairly self-explanatory, and you probably discovered your favourite book this way – just picking up a random, interesting looking spine at the bookstore and taking a chance with it. The method of searching for your favourite genre and finding a book is also considered ‘serendipitous’, though that doesn’t sound as cool.

Both social and distributed book discoveries come under the same umbrella: word of mouth. Friends, social media and big media are responsible for these types of book discovery, whether it be a recommendation dropped in conversation or a newspaper review in the Sunday edition. Social media and the Internet revolutionized the ‘social’ method, allowing people to share their recommendations to hundreds of people at once and extend a branch of friend-readership for books and readers that might not have been brought together otherwise.

Amazon, being one of the foremost ebook sellers on the Internet, would be the obvious choice for ‘best book recommender’ under the category of ‘data-driven discovery’. They already utilize a system that analyses user inputs like previous purchase history, user ratings and reviews to make recommendations for like-minded customers (‘If you like X, maybe you’d be interested in buying Y like this other guy did!’). These recommendation systems can be found all over the Internet, from music to Youtube to clothes to academic articles. For more ways on how Amazon recommends things, check this out.

Incentivised discovery is, again, similar to serendipitous discovery, but this way is better… because it’s free. Promotional material, review copies and giveaways provide readers with the opportunity to discover books they might not have otherwise. People will take all kinds of stuff if it’s free.

In the future though, apparently most of these methods just won’t (or maybe shouldn’t?) cut the mustard. According to some sources and writers on the Internet, Google is the best possible source to receive book information and recommendations from. If you think about the amount of personal information that Google has access to, then it makes sense. It has access to four of the types of book discovery.  Amazingly enough, if you use the Google Play Store to purchase apps and access and read ebooks, then Google already knows what you like to read and how much of it you read.

Liquid State - Google Book Recommendations

If Google’s already into all of these, why not get into the recommending game, too?

Furthermore, if you’ve ever used Google Search, Google+, Gmail, or Google Chrome (which I’m betting you have), then they also know what you search for online, what you chat about with your friends online, what books you email about or what book-related enewsletters you receive, and what you read online. Phew. I don’t think they’ve really missed anything – though Google can try and create serendipitous discovery of books, they really are shooting in the dark without using information gained from the various Google programs available today.

Just to sidetrack for a moment – the blog discussing Google’s future dominance as a book recommendation engine also claims that discussing books in a brick-and-mortar shop is far easier than online, and that hundreds of books can be viewed in minutes in a brick shop while only 20 books can be viewed at a time online. This annoys me, and I think it’s a generalization that reveals the authors obvious bias towards physical bookshops. Sure, you can look at hundreds of book spines, but I would be very surprised and impressed indeed if the author of that blog managed to read the blurb and get a good overview of what the book is about in the same time. If all you’re doing while Web surfing is scrolling and clicking to find books, surely that would take just as much time, if not less.

I just couldn’t help myself there – let’s carry on. If Google realises this potential and decides to act on it, there could be some very interesting, big implications for book discovery, and for book behemoths like Amazon. I’ll admit, when I read that blog explaining about Google I had to sit back and think about it for a few minutes, because it certainly is a big idea, and would require Google to change it’s current business model to one more suited for a consumer-oriented retail company. Of course, I’m not saying that this is inevitably going to happen, but it might someday, if Google or someone like Google decides to take on the challenge.