Comments on: Two years and three CEOs later, publisher JV Bookish is ready to help users find their next book

By: Jenny Glover

Jenny Glover — Mon, 01 Apr 2013 12:58:47 +0000

I have bought a book supplied by Booki.sh. (a) because it was over$5 cheaper than others. (b) because it claimed to be capable of downloading to my new Kindle 3rd.Generation Wifi and 3G.
Currently I am waiting on a reply as to why the download starts but stops after about 5% complete. I did however, have no problem downloading the book to my Android tablet.

Booki.sh will need to fix this glitch with Kindle downloads quickly to avoid negative feedback to consumers and Retail bookstores.

As to comments about Cloud reading and not having the opportunity to a fileable function. If the price is right, I think there lots of buyers, including myself, who couldn’t care less as long as we get to read the book and it remains available. Cheers Jenny

By: Adam Gobrat

Adam Gobrat — Mon, 18 Mar 2013 05:36:33 +0000

great discussion. apologies for joining late, has anyone heard of boikeno.com. they seem to be focussing on ebook only though. but i agree discovery is a great problem that needs to be solved.

By: Andrew Chapman

Andrew Chapman — Thu, 28 Feb 2013 07:37:24 +0000

Any algorithm will throw up odd results sometimes, and at least that allows for a bit of random serendipity. I co-run http://www.whatshouldireadnext.com, which has been running for 7 years now. Our system is another collaborative filtering one predicated on users’ taste rather than buying or reviewing patterns. Content-based or metadata systems have their place too, but can get bogged down by detail (others have come and gone in this space). Bookish looks nice, though it’s hard to see what’s new there, other than publishers coming late to the discoverability table. It feels a bit like Anobii, maybe. Goodreads does the social side, LibraryThing appeals to book obsessives, our site offers quick and snappy suggestions without any frills, and appeals particularly to young adult readers. Take your pick! Maybe we need a whichdiscoveryengineshoulditrynext site…

By: Angus Swan

Angus Swan — Wed, 13 Feb 2013 16:02:39 +0000

Putting the recommendation technical challenge to one side, the major commercial issue with book recommendations is how to monetize them.They are so prone to ‘showrooming’. So – you think I should read this? Okay, i’ll just go and buy it wherever i can find it cheapest..You can’t charge for the recommendations service, but you can’t couple the service to the purchase. This is not a new problem, and it should not put publishers off creating services like these as long as they regard them as marketing costs and not revenue opportunities.

By: sorrygnat

sorrygnat — Wed, 06 Feb 2013 23:25:06 +0000

i don’t know; i always like to support such endeavors. I use Alibris when it comes to used books and don’t have the income for books on line. I go to the library or independent bookstore Vroman’s; but i’m open to it all. I teach the Courage to Write, and am happy to recommend it; where can I go for initial language and Pr STUFF?

By: Alexander Braun

Alexander Braun — Wed, 06 Feb 2013 11:43:24 +0000

@David Thomas: True, the algorithm can only be as good as the metadata it’s relying on. With http://www.semantopic.net/ we’re covering 1.2 million books with more than 110k tags/categories, giving recommendations based on topical book similarities a much broader base.

By: Laura Hazard Owen

Laura Hazard Owen — Tue, 05 Feb 2013 20:03:43 +0000

@andrewlos You’re absolutely right — Khazaei just confirmed this to me, and I’ve updated the post (see above).

[Updating my own comment]: It’s a tricky balance…steering clear of DOJ, avoiding angering retailers, and still selling direct — tough combo. I realize now that Khazaei can’t have a lot of direct knowledge of it, but it does seem as if the individual publishers will now have the chance to experiment with some discounting on their own, individually.

By: andrewlos

andrewlos — Tue, 05 Feb 2013 19:33:16 +0000

I’m not sure they can get more directly involved without falling foul of anti-trust law, due to the publishers behind their company. They represent the interests of too high a percentage of the industry – if they started to set prices, the DOJ likely get interested again.

By: David Thomas

David Thomas — Tue, 05 Feb 2013 19:20:25 +0000

I think the general inference that what book or sequence of books purchased is indicative of what a consumer wants next is far too simplistic. Netflix implores users to rate what they watch for example — and they can evaluate what was abandoned, the relative rates viewing series, etc. I think the data culled from Hiptype would be far more useful for Bookish to evaluate the arc of a reader’s interests than sales events, social media “likes”, and the Good Reads bookshelf. Myself, with Good Reads, I have yet to see a recommended title that is actually on my projected reading list.

By: Eric Welch

Eric Welch — Tue, 05 Feb 2013 17:25:06 +0000

I’m a very heavy user of Goodreads (almost 1300 reviews, some 1500 books read) and I’ve been playing around with the Goodreads recommendations engine which makes recommendations based on the user’s shelf assignments, e.g. suggestions based on what I have read or want to read on my biography shelf. Interestingly, I rarely look at those recommendations but use the purchase suggestions on Amazon a LOT. I buy many books, almost exclusively ebooks now, perhaps $300-$400 per month, and find those suggestions invaluable for locating new different books on a particular topic based on what other customers viewed and purchased. It really works pretty well. To succeed Bookish will have to find something they can do better than the long-established competition. They have a long way to go.