Finding Really Good Books

Finding really good books is a multi-billion dollar market. Think about it for a moment: How do you find new books to read?

I assume it will either be through:

• a recommendation of a real person you know (the best way),
• a recommendation you read online (including Goodreads),
• an algorithm,
• an advertisement,
• or a bestseller list.

The best way to find a new good book is through a recommendation. You are conversing with a real person, he understands your situation, your needs, or your preferences and based on this, he is recommending you a book.

Another a good way to find a new good book is through a recommendation you stumble upon online. You search for your preference or a specific topic or problem, and you will find online communities or blog posts pointing you to a specific book.

Algorithms, advertising, and bestseller lists all have one thing in common: they want to sell you something. Despite that, algorithms can still dig out good books for you, based on your preferences and past reading history. But don’t expect to ever find real exotic books outside your current interest bubble.

Books on bestseller lists can be good, but generally, they are simply the best marketed books or have the best known authors.

Again: bestseller lists are listing the best-selling books, not the best books.

The question is therefore, where do we find, and how do we identify the best books? And why is it an important problem we should solve?

4 Million Books

Let’s say on a bestseller list you can discover the 20 best-selling books on a specific market. On the New York Times bestseller list, you’ll find the 20 best-selling books in the U.S. On the SPIEGEL bestseller list, you’ll find the 20 best-selling books in Germany.

Let’s say we add multiple markets together. With many overlaps and translations, we’ll optimistically end up with 100 best-selling books in all Western markets.

That’s only 100 out of roughly 4 Million books published each year worldwide – including self-published books. That’s roughly 0.0025% of all books published.

As we’ve said, the bestsellers are listing the best-selling books – not the best books.

With a very high certainty, there are several thousands of books which are better than those on the bestseller list, but whose author simply does not know or does not want to market the book heavily and professionally.

A lack of marketing and selling, however, does not make a great book a bad book. It only makes it an economic failure. But it is still a great book.

Recommending the Known

When you think about the last time someone recommended you a book, the probability is quite high, that he found this book on a bestseller list, through advertising, or an algorithm. This means, we are ending up in the same recommendation and feedback loop: the best marketed and best-selling books get recommended over and over again – while some of the best books ever written in human history might remain undiscovered.

The Billion Dollar Market

Finding a really great book is therefore a billion dollar market. Why? Because among 4 million partly self-published books, the majority will be bad (really, really bad). People will naturally stick to books directly recommended to them through friends, other individuals online, an algorithm, advertising, or a bestseller list. Because it is today the best shit-filter we have.

We need a new technology to discover the diamonds among several millions of published books. And this – ladies and gentlemen – is a multi-billion dollar market.

What may be a solution? I don’t know the solution, but here’s what comes to my mind.

People get served pieces (abstracts, chapters, etc.) of an anonymous book. After reading it, they rate it. When they read the piece of the book, they don’t know the popularity, they don’t know the price, the book cover, the author, or the ratings of other readers. They rate the book totally unbiased. Based on these ratings, one could create non-commercial book recommendation algorithms or deep learning models.

Only by making everything anonymous can we judge a good by its quality – not by its cover (or bestseller rank, advertising budget, etcetera).

Whether this is the best way to do it? I don’t know. But finding really good books is a problem worth to be solved.