IfBookThen: Letters by Numbers

by Suw on March 21, 2013

Tove Leffler, The Swedish Bookseller

Discoverability: how do we find books and how do we find readers? Over last ten years, number of books in bookstores has decreased, see less books and more other stuff like cards and toys.

Death or the physical bookstore that has happened in UK and US is not yet here, in Sweden bookstores have grown since 2005. Per capita, Sweden has more bookstores than US/UK. But, in small decrease, 3% have closed in last few years.

In 2011, 75% of bookstores decreased turnover. More stores will likely have to close.

Don’t buy as many books in bookstores as we used to. Apart from that being a problem with revenue. There is another issue to face: Where do people find their books.

Article in Swedish Bookseller about struggling literary reviews. Reviewers are paid less, fewer books are being reviews, newspapers facing huge losses in readers, so reviews not necessarily a good way to reach readers, esp young ones.

Internet can be  useful, all familiar with concepts of ‘also bought’ algorithm. Easy to search for a book that’s been recommended. But what we lose with the internet is serendipity. Hard to find things we don’t now we want or have never heard about, or outside our comfort zone. That is something that online bookstores have to figure out. Because it’s not really about randomness, it’s about being told about something you would never think of, or that people who have the same taste as you would never think of.

Bookish is trying to work around that with its recommendations from staff, and make recommendations from a more complex algorithm than other systems. But that’s hard, because when you come to computers someone has to do the coding and you lose a bit of ‘human error’, and that error is vital to finding new reading, or new readers.

Need to understand readers. Eg. Where are they? What else to they buy? What do they read? Why do they read? What do they eat? What music do they listen to?

Finding this out is easier on the internet.

“If you like this book, you will most certainly not like this book.” Good to recommend the unthinkable.

Nick Sidwell, Guardian

Guardian Books is a very small publisher, trying to do something that is different, have access to the wider environment of the Guardian newspaper.

Have access to a vast range of different writers, and to an enormous audience. 30m monthly unique browsers, 215k print circ, global audience, millions of followers on social media, 10,000 of bookshop customer.

Useful to tap into this audience, but these figures are the equivalent of sales figure in that they are straightforward measurements of a single factor, but what’s more important is the data behind these figures.

Based on what was popular amongst the paper’s reader, decided to publish a book called Swim, which is about swimming in wild places.

Now trying to construct more of a story from the data.

Guardian Shorts, ebook only short-form non-fiction or long-form journalism. Same thing. No longer just taking popularity factors, but also asking questions of that data and understandings of the audience. So understand how they engage with content, how they behave on the website. Team of data analysts take from that deeper, richer more targeted and more focused understandigs of what it is that people want. Then apply that to editorial decision making.

Three questions to ask ourselves:

1. Which subject areas should we focus on?

2. What should that content be?

3. How big is the potential audience, who are they and where can we reach them?

Have a lot of data about who these people are, where they are, what devices they have. Can answer some of the questions Tove mentioned in the intro.

Most popular areas on news website: news, sport, culture, travel, tech. So develop shorts in those areas.

Now have nearly 60 titles. Take a lot of archival content from both the Guardian and Observer, and with editorial work can curate that into short packaged ebooks.

Now trying to move beyond curating archive content, into commission new stories, not just a title or subject, but to use that extra data for how we structure of the book itself.

Facts are Sacred is about data journalism. Used how people were engaging with the Guardian data blog, knew there was a very engaged audience and looked at what they were interested in, and looked at what was keeping them coming back, what was important. Decided on eight chapters to understand what was done, whether they could do it themselves, to how it changes the role of journalism.

Simon Rogers, editor of data blog, wrote the book. Were able to use understanding of the engagement that people were showing to predict what their possible sales were. Knew that 10% of 1mn audience were engaged, knew how any had ereaders, how mmany in UK or US, and how many read short-form non-fiection — 4,844. Also knew from archive content, and knew 50% of sales came from non-core Guardian readers, so realistic figure for sales target is 9,699. Have sold 7,000 copies so far, but that ebook is no longer available as redeveloping it.

Felt happy with that, was a very useful proof of what we were trying demonstrate. Could take this deeper analysis of how people respond to the newspaper to inform decision making for book content.

New editions published next month, moving to a printed book an a rich ebook developed for the iPad, which is better for the type of content, including videos and interactive elements.

Been using marketing data to try and match where we’ve got to with our editorial decision, and then to make sure that we bring it to the right people, commercial profiling of customers, so can match up characteristics of book buyers who have shown engagement with similar products.

What lies at the heart of this is the editor. It’s still the editor that makes the decisions, the data without someone to make use of it is just a big spreadsheet of numbers. Even when the different metrics are put together to generate insight, unless we know what to do with it, ask it specific questions, have a goal, it remains a spreadsheet full of numbers. Editor remains central.

Still need all traditional editorial talent. Don’t let the data dictate what we’re doing, use it to inform decision making and understanding. It’s a tool much as an editor will take on all sorts of information about a market, the data about our reades is just another tool. Allows us to make wiser, more informed decisions.

Been doing this a year and a half. At the beginning of this year we made a switch, moving away from the archive material, commissioning more new titles. All of those titles have come from an understanding of our audience.

Great opportunity to gather more data. Want to not just use data to commission, but also to develop how we distribute. If you understand how people read your books, that’s possible digitally, not always easy to get hold of, but if you have channels that feed back to you, it can be enormously valuable.

– be data first

– data is a tool editors shouldn’t be without

– data needs organising and interpreting

– use data to prove your assumptions

– …but also allow it to change your mind.

Andrew Rhomberg, JellyBooks

Industry gorilla – Amazon. How do we not get squashed? Where is that gorilla weak? Discovery. Amazon is where you go to buy book, few peopel discover new books on Amazon. Amazon doesn’t share data. If we have a data-focused approach, can we use that data and collaborate around it? Can we be DRM free? Can we share book samples?

Discovery is not one thing. Five forms of discovery:

– Serendipitous

– Social

– Distributed

– Data-driven

– Incentivised

Jellybooks.com had just covers, no price, no text, because it’s easier to browse pictures.

Want to make books more viral and engaging, so create a ‘twitter card’ which is easy to share.

Widgets for book sampkes so authors, agents, publishers and reviwers can embed them on homepage blog or website. Samples which are easy to include, and people can download a sample later for reading.

Oldest data-driven discovery is the best-seller list. Can have a data-driven approach that’s unique to each person, not just mass popularity. In a wider concept, think of book as the paper book, butthinkn of the ebook as a file, a container. Couldthink of a book as a URL, maybe to the produce page, Goodreads review, quote that has been pulled out on readmill, these links are shared over the internet – blogs, Pinterest, emails. These links are accessible, can’t see amazon sales data but can see how People share link to Amazon product pages.

Who’s sharing? Who’s acting on that sharing? Use a book link as a reference, and because have index the links, can track them.

Data is very messy, so track it, clean it up, and put it out as an open API, so that others out there can use that data and create fabulous new ways of discovery. People can do data visualisation, data mining, and others benefit from their work.

Jellybooks data: want to know what people are reading, without asking them or them telling us. If you ask someone what they are reading they will not tell you the truth. Books on the shelves are what they want you to believe they are reading.

So through the links on sharing and consumption can get an idea of what they are really doing.

If we want to influence a reader, can you influence the people who influence them rather than spam them?

Any publisher can participate, and then they start tracking the books. Discovery only of interest when it’s their books being discovered.

Incentivised discovery – you get a special discount but only if you can get enough people to join in. Group deal. So get an email, download the sample over breakfast, first 10% of book, share it with others, and if you like the deal you sign up with your credit card. But you need to share it, and get others to buy it as well.Can monitor how you are progressing, don’t tell you what the minimum is (set by publisher) the deal is activated and you get the book discounted by 50%.

Deadline is 6pm, so there’s only 12 hours to do it. Most people in London commute home at about 6pm, so they can then get the whole book to read on homewards commute. Change those times in different countries, eg. Spain is 9am – 9pm.

Not about trying to kill Amazon, but about an alternate strategy.

Soon launching in Spain, US, Latin America.

May be targeting a smaller market than amazon, but because it’s such a big world, can make this niche big enough to be viable.

Comments on this entry are closed.

Previous post:

Next post: