Digital Book World E-Book Best-Seller List Methodology

Expert publishing blog opinions are solely those of the blogger and not necessarily endorsed by DBW.

When we decided to build our new Digital Book World E-Book Best-Seller List, we wanted to give people a better view into what is actually happening in the e-book marketplace than they already had.

In the absence of real-time, raw sales numbers for all the retailers — which nobody has — we thought it best to address some of the problems with current e-book best-seller lists:

— One-hit wonders
— No differentiation between $0.99 e-books and $12.99 e-books
— No accounting for which publishers are selling the most e-books

In response, our list measures sales ranking over a week’s time to reward books that are best-seller lists every day of a week versus just one day. We also show price and have created four separate lists that organize all the books into price categories. And we show the publisher, so the industry can keep track of which publishers are selling the most e-books and how they’re doing it.

Below is basically what we do every week. Unfortunately, we won’t be exposing our full methodology as we don’t want to give everything away but the below should be telling. Readers should also know that certain industry experts outside of Digital Book World and Iobyte Solutions (my company, which is powering the raw data behind list) have taken a look under the hood, so to speak, and help influence our methods.

Those experts are listed in the press release.

Our basic methodology:

1) Best-seller rank observed from the six top retailers (Kindle, Apple, Nook, Google, Kobo, Sony)
2) Lists observed for seven consecutive days (Sun. – Sat.)*
3) Each appearance on a list gets an unweighted score based on the ranking
4) Ranking scores are logarithmically determined (i.e. top scores are much more valuable than lower scores)
5) Each retailer weighted by approximate market share as determined by the editors of Digital Book World and Iobyte Solutions
6) Additional appearance credit is awarded for appearing on multiple lists
7) Combined scores for the week determine final score for each title
8) Titles are ranked by final scores and also grouped into sub-lists by price (four separate price-band lists: $0 – $2.99; $3.00 – $7.99; $8.00 – $9.99; and $10.00 and above)
9) Minimum price that appeared at any point during the week on any retailer is used for determine price band (assumption that low price is an important driver of ranking)

Let us know your thoughts: Leave a comment below.

* Note. This has been changed as of the week of 2/3. See editorial note here.

17 thoughts on “Digital Book World E-Book Best-Seller List Methodology

  1. Dan Lubart

    Stacey, Not right away, but it’s something we have talked about. We’d like to let these initial lists bake for a while and then see what the market wants next. One thing I’ll add is that I actually like the idea of a combined list where you can see how \The Hunger Games\ compares to \Fifty Shades of Grey\.

  2. Jon

    Are there plans to include books sold in the Apple App Store and/or iBooks?

    I also vote for separating fiction and non-fiction. You may also want to consider separating $0.00 from $0.01-$2.99.

  3. Michael

    Not sure how tracking rankings can give us the most accurate view, cross-retailer. Do we know, for instance, that each retailer uses the same methodology when determining these rankings (i.e., units vs. revenue). Additionally, each retailer updates their rankings at various frequencies (Amazon is hourly). Do you track these rankings at the same time each day? Also, banding together free and paid seems problematic as these are tracked and ranked separately for most retailers, creating two separate data sets, and thus, requirements for where a title lands on its respective lists. If you then combine these rankings and weight them equally, without access to the raw data, you have no way of really knowing how they compare to one another.

    I think this is an admirable effort, and in the absence of cooperative sales reporting by the retailers may be our best available guess. I would just caution that this is not exact and is still only an educated guess at what a combined bestseller list looks like.

    1. Dan Lubart

      Mike, a few quick answers.

      1) I think it’s a certainty that the retailers use different metholdogies for their rankings. I have compared to real POS data enough to know that rankings are generally a reasonable proxy for unit sales, although FAR from perfect. I track once per day late at night so as to best avoid any media-driven spikes that could impact hourly rankings. It’s a best guess on my part as I am not privy to the retailer ranking methodologies.

      2) I don’t expect to see free books on these lists since I track the paid Kindle list for this purpose. I agree that combining free and paid is problematic. I can’t control if a free book slips into any of their lists and I am not looking to filter anything out but I wouldn’t expect it to be a significant issue.

      3) I heartily agree with your last paragraph. This is exactly an educated guess. There just isn’t a consistent, timely data source available to do better with that we know of. I compile this data and analysis already and some colleagues mentioned that an aggregate list would be interesting and so it was born. I hope over the next few weeks, as we tune things in response to our observations and those shared by others, we arrive at something generally useful and interesting.

      Thanks for your comments.


  4. Thomas Ellsworth

    As a data junkie myself and proponent of analytic-based business assessment, I applaud the methodology itself and attempt to deliver a highly usable industry report. Clearly both of the foregoing will evolve over time and it is my hope that this progress influences the storefronts themselves (well, I can hope, can’t I?).

    Full disclosure – I am not part of DBW or F+W. I am CEO of Premier Digital Publishing – thus my keen interest in a regular, reliable source of bestseller sales information. My career started in marketing before detailed data as collection and the tools to crunch it emerged. Today, those who know me have heard my “marketing is now math” speech. I am quite pleased to see that approach come to the analysis of eBook sales and congratulate DBW and Iobyte for taking the initiative to do it.

    Well done, guys.

    1. Dan Lubart

      Sincere thanks, Thomas. Very well put. We will strive for perfection, knowing we’ll never achieve it but do our damndest to get as close as we can. Along the way, I’ll try to share more about the methodology employed, including when and why we make adjustments. We’ve already gotten some great feedback from posters in a few places and will address a few of them soon. I went into this knowing that almost every decision we make and publicize may be viewed as arbitrary by some, and open to criticism from many directions but that should only help make it better in the long run.

  5. Frank Daley

    The book industry is notorious re secrecy for total sales, book amounts earned, best sellers which definition would you like to use today?!), author earnings, etc. Good work in a dense field!

  6. Bill Peschel

    Interesting effort. It looks like it’s capable of clarifying what’s happening out there in the market.

    Some suggestions: 1) Separate the free books. Seems like an unfair competition, even with the $2.99 price point. 2) Separating fiction from non-fiction would be good, but from a look at Amazon’s nonfiction bestseller list, I see a lot of fiction there (even if they are tagged as nonfiction), so many that I wonder if that’s not a scam technique to get onto more best-sellers list. 3) Would love to see a top 20 or 25 instead of a top ten, to get a better idea of what genres are preferred.

  7. William Ockham

    The logic behind step 6 is highly questionable. The difference in the market share between the top and bottom retailers is so vast that it seems a bit nonsensical to me. How many authors would trade being, say, 26th on the Amazon best seller list for being first on Sony, Kobo, and Google, but not being sold by Amazon?

    1. dan lubart


      I think a more accurate comparison is trading being 22nd on Kindle only for being 26 on Kindle but also top-20 on those other 3 lists. Differences in sales units may be very small in that case (numbers are illustrative only, not meant to be reflective of real data) but we choose to slightly favor the books that hit multiple lists in that case.


  8. Peter D Sowatskey

    Good information.
    By looking up each book I could find out the content. I saw the observation that erotic romances are faring well now.
    Whta’s popular is very important for me. I write a book every three months; that’s from start to download on Kindle. I can easily change my emphasis in a story line to whatever’s selling now.

    Perhaps you could add another colume to show the basic theme of different books.

    Thanks, Peter

  9. Lisa Grace

    Amazon weights a free book approximately 1/10th that of a paid book. Self-published books appear to not be updated (ranking wise) as quickly (approximately a six hour lag), but seem to go down faster. Self- published books also do not appear in as many categories (only allowed to pick two when uploaded) for best seller list purposes, thus diminishing their ability to stay at the top as long by simply not being as visible.

    Amazon far outsells any competitor and I don’t know many authors who would trade being #1 on any list over being in the top 50 at Amazon ( for a seven day period.)

    What about Wool? I find it hard to believe its #1 on so many lists and didn’t even make your top 25. What about foreign (like UK, CA, DE, ES, etc.) Amazon sales? Are those factored in too?

    I think your methodology is not quite where it should be, yet. (Just my personal opinion.)

    1. Dan Lubart


      Thanks for your comments. I can’t speak for authors in general as to whether they value a higher rank on Kindle more than being #1 on other lists. For some, it’s just about sales, but for others, there may be other factors in play. We did speak to a number of authors and agents and came up with what we felt was a reasonable position, hopefully ideal to some, suitable to many, and probably not so suitable to a few. I don’t know that, in the absence of real sales numbers, there is a ‘better’ solution but we’re always talking about it and making small adjustments as we can.

      We don’t weight free books at all right now. Perhaps that’s something to consider and I’m open to hearing more on that but for now, our focus is on books that people are willing to pay for. As for Wool, it’s currently ranked 32 on Kindle and 157 on Nook, but there are occasionally issues that affect a specific title because retailers don’t all use the same identifiers. I will certainly check and see if adjustments can and should be made and get back on that.




Your email address will not be published. Required fields are marked *