Previous month:
September 2016
Next month:
November 2016

2 posts from October 2016

Here’s a better model for book search and discovery

Screen Shot 2016-10-15 at 3.49.38 PMHow are you helping consumers find the perfect book for their needs or interests? If you’re like most publishers, you offer a search function on your site. Visitors simply type in a topic and relevant titles from your catalog are displayed.

This is pretty similar to how search works on Amazon. In both cases, book metadata is used to determine the best matches. So if the search phrase happens to be in a book’s title, description, etc., that title is likely to float to the top of the results.

That’s great, but why not leverage the book contents, not simply its metadata, for the search process. Amazon’s Search Inside feature lets you do this, but only after you’ve selected a particular book. What if you’re a publisher with a deep catalog on religion and someone is looking for the book with the most in-depth coverage of Pope Francis? Metadata-only searches can help, but the full contents are the only way to truly measure topical depth, especially if you want to compare two similar titles to see which one has the most extensive coverage of the search phrase.

Google Book Search (GBS) offers this sort of visibility but most publishers have a cap on the percentage of content visible to GBS users. That’s primarily because publishers want to prevent someone from reading the entire book without buying it.

I believe the solution is to expose all the contents to a search tool and display results that only show snippets, not full pages. That’s exactly what we’re now offering on our bookstore website at Our Sunday Visitor. If you click on the Power Search link at the top of the page you’ll be taken to this new search tool.

If I search for “Pope Francis” I get these results. The top title has 203 hits, so if I click “view 203 results” I can then take a close look at every occurrence of my search phrase in the highest ranked title. Note that this platform takes proximity into consideration, so if you have a multi-word search you can limit the results to just those instances where the words are closest to each other. At any point the user can click on the cover image to read title details or buy the book.

Think about how powerful this tool is for publishers with deep lists on vertical topics (e.g., cooking, math, science, self-help, etc.). Instead of relying exclusively on the book description to make the sale, the contents are fully searchable and comparable across a list of related titles.

We’re in the early experimentation phase with this platform. We’re planning to use a variety of ads that say something like, “find your next great read”; users who click on those ads will be taken to the search landing page where they can explore the full contents of our entire ebook catalog.

This search platform is powered by the outstanding team at MarpX. If you’d like to experiment with this on your site, you’ll find contact info at the bottom of their home page. MarpX has been a wonderful partner for us and I highly recommend you explore their solution as well.

I hope you’ll join us in this effort to move content search and discovery to the next level.


Google experiments with book discovery…and fails

IMG_0008Even though you probably never stray from the Kindle reader app I’d like to encourage you to expand your horizons. It’s a good idea to keep an eye on Apple’s iBooks and Google Play, for example, to explore other platforms and keep Amazon honest. After all, Amazon’s need to innovate diminishes if ebook platform competition dries up.

When Google recently announced plans to add a Discover feature to their ebook reader app I was curious to learn more. Google is the king of search so I was hoping they could use their brawn and data to create a major breakthrough on the book discovery front.

I assumed Google would look at my Play ebook library and base some assumptions on what I’ve bought and read over the years. I figured they’d let me recalibrate their assumptions to better suit my interests; for example, they know I like hockey books but my Google purchases haven’t focused on my favorite team, the Pittsburgh Penguins. Lastly, since Google monitors my Gmail inbox and search requests, I also assumed they’d use that info to fine tune their book recommendations in their new Discover service.

My hopes were dashed and my assumptions proven wrong when I saw the results. Google Discover is nothing more than a dumping ground of all things books. They apparently assume that if you read books you’re interested in everything about books; that’s like assuming a 70’s rock enthusiast is interested in all types of music including disco, jazz, classical, rap, etc.

How could Google get it so wrong? Why did they simply mail it in and why did they even bother? I’ve got to believe usage of Google Discover is pathetically low. If so, I hope the poor performance doesn’t discourage Google from going back and doing it right the next time.

Google needs to leverage all that data they have about us, more than Amazon has, btw, go back to the drawing board and come back with a Discover 2.0 service that really works and is deeply engaging.