Ranking e-book sales

Last week The Bookseller announced that it would begin running monthly e-book rankings, with data supplied by all the main trade publishers. The initiative builds on what we've done with The Bookseller's annual, and quarterly market analyses, where we've integrated publisher supplied e-book data with stats from the physical book market supplied by Nielsen BookScan.

We first began having a conversation with publishers in December last year, as I felt that attempting to analyse the publishing market only through the prism of print book sales was becoming untenable. It is worth remembering that BookScan has existed for less than 20 years, yet we have all come to rely upon it for timely and accurate information about what books are being bought and sold in high street bookshops, supermarkets and the internet.




BookTrack, as it was first known, was founded by Whitaker, which at the time also owned The Bookseller. It was a great innovation, and continues to be so, operating now in 10 countries, having most recently announced a new panel in Brazil.

The Bookseller took its first BookTrack-supplied chart in the final issue of 1996. Its predecessor was a service known as Bookwatch, which provided estimated sales figures (and rankings) based on the reports from a small panel of bookshops. The Bookseller also used to run an annual chart known as the 'Fastsellers', compiled by Alex Hamilton, and based on publisher supplied sales figures of titles published in the relevant year.

The history lesson is important. Bookwatch provided charts to The Bookseller for close to 20 years, but could only offer estimates, as opposed to BookTrack's solution that took sales data direct from the tills operated by the participating booksellers. The step-change in our knowledge of the marketplace was huge. Shortly before it switched from Bookwatch to BookTrack, The Bookseller prepared the groundwork stating that "except for anecdotal data of variable reliability, we have little evidence of anything at all about the precise size or behaviour of our market".


Digital sales


The irony is, of course, that speed forward 17 years, and those words could just as well apply to the situation we now have around digital sales. The e-book market has grown to near 30% some trade publishers' business but there is no third party monitor of it, or its "behaviour". In the US, where the e-book market is bigger, the data sources are no better. The New York Times and the Wall Street Journal both run e-book charts (the latter supplied by Nielsen BookScan, and based on sales from Amazon.com, Barnes and Noble, Apple, Google, Sony and more with Kobo joining the panel shortly), but without numbers. Digital Book World's weekly e-book bestseller charts are based on sales rankings at e-book retail sites, weighted for market share and a number of other variables (similar to the e-book charts first run on FutureBook). The Associaton of American Publishers also produces monthly data about US publishers' sales known as Stat Shot, which is then rolled up into an annual BookStats report, done with the Book Industry Study Group.

Yet nearly everything we know about the e-book world is based on publishers' financial statements, analyses of relative movements of titles on various e-book bestseller charts, or consumer surveys. We know something, but not enough, and certainly not enough for a publishing community that has grown up with an abundance of data as supplied by BookScan.

The monthly e-book ranking

So The Bookseller will begin running a monthly ranking with publisher-supplied volume data with the first ranking to run on Friday (2nd August), based on June data. This will initially begin as a top 50 all titles chart. The data is aimed at matching Nielsen BookScan data from the print market, so we'll focus on UK sales (i.e. Great Britain and Northern Ireland).

As with the BookScan charts we'll focus on volume, but will eliminate titles that are so heavily discounted that their sales figures become meaningless. The Bookseller's BookScan charts exclude titles that have an actual selling price that represents a 75% discount or more on the rrp. There is no clear way of making this calculation for e-book sales, so in the e-book ranking we'll exclude titles that sell below £2. This is consistent not only with our own existing charts, but also with what other e-book charts there are out there, including the Wall Street Journal's, which excludes free e-books and e-books sold under 99 cents.

The publishers in the launch panel are Penguin Random House, Hachette, HarperCollins, Pan Macmillan Bloomsbury, and Simon & Schuster. All have previously supplied e-book sales data to our annual and quarterly reviews, so we are confident both that they have the systems in place to interrogate the data they receive from retailers, but also that they can get it to us each month at around the same time. Beyond these founding six, we expect to grow the panel quickly: so far we've added Canongate, Faber, Hesperus, and Short Books, and more will follow for the July chart. The first six publishers represent 51% of the printed book market and 81% of the print fiction sector.

For guidance, we've run one trial chart with these six publishers, with the highest ranking book on the chart selling circa 150k, and the book in 50th position selling 4,500 copies. The chart was dominated by fiction, with only one non-fiction title making the list.


A different perspective on the e-book


Aside from showcasing the biggest selling e-books in the month, the ranking will also grant us a different perspective on the e-book market (as I partly alluded to here). So far, we've looked at e-book sales in relation to their print sales, but the e-book market is growing up and moving swiftly in a different direction. Our trial chart showed an E to P+E ratio of sales greater than the overall figure of 30%. More than a dozen of the titles that featured in that ranking were selling more digitally than in print. This market is different, and it needs to be measured directly. Though we'll initially focus on the Top 50, the ranking will allow us to track sales growth over time, and make estimates about the wider digital market. Both Bowker and Nielsen BookScan already do this via their consumer panels, and it is to be hoped that this monthly sales data can help to underpin this good work.

Neither the ranking nor the approach is perfect. There is no third-party monitoring of the publisher supplied statistics, and though we will be picking up publishers for anomalies (we continue to track Amazon's Kindle charts for the relative movements of individual titles, and prices), we will be reliant on their honesty when supplying the data. Unlike the old days of the 'Fastseller' chart, it should be relatively straight-forward to pick up the odd rogue title, based on how we've seen it performing on Amazon's Kindle chart.


Digital discount


Though we are emulating the discount cut-off threshold when it comes to eliminating cheap and free titles, I'm not sure the two situations are wholly comparable, and at £2, my concern would be that genuine bestsellers will not feature in the ranking. As 20p showed, price is huge mover of volume sales in the e-book world, much more so even than in print, and we will need to keep an eye on this threshold to make sure we are able to reflect this. I favour a standalone ranking, but some publishers have suggested running charts that are segmented by price bands, as DBW does. We'll continue to reflect on that.

Last the panel of contributing publishers will never be wide enough to satisfy everyone. We'll look to include all publishers we can do, including over-time self-publishers and e-book only companies, though we'll need to be careful each time to make sure their data is as robust as that of the big publishers (we've already begun talking to the Alliance of Independent Authors about whether/how we can begin to collate indie author e-book sales).

Nevertheless, getting publisher specific data is different from getting it from retailers, and it is for this reason that we are calling the chart a 'ranking' rather than a bestseller chart. This is a ranking of e-books based on data supplied to us by participating publishers. As with the Bookwatch charts, the ranking may in the end be superceded by a more robust solution ideally with sales data supplied and made publicly available by Amazon, along with data from the other smaller e-book retailers. If Nielsen can bring to the UK, the process it has already begun in the US, all the better.

The ranking shows how 17 years on from BookTrack, UK publishing can still innovate. We will be the first place to introduce an e-book ranking with publicly visible sales data anywhere in the world. I'm grateful to the publishers I spoke to in the planning of this, who were generous with their time, and positive with their feedback. No-one said no. And every publisher I spoke to wanted the same thing: an understanding of the new frontier. This begins that, but it by no means the end of the process.

Read more on FutureBook