• Print

Neat visualization of download ratios for ebook formats offered by O'Reilly

PDF has competition from EPUB.

At O’Reilly we offer multiple (DRM-free) formats to choose among for customers who buy our ebooks. Since starting the program with PDF, EPUB, and Kindle-compatible Mobipocket formats, we’ve added an Android application file (.apk) and more recently the accessible DAISY format.

We track which of the formats customers actually download, and from the start PDF has been the dominant choice, though as the chart below shows, there’s been a steady shift toward other formats, especially the open EPUB format (which can be read on nearly every ereader device, and is the format used by Apple’s iBooks reader).

Screen shot 2010-09-30 at 5.16.35 PM.png

In some cases, customers download multiple formats, but this data just includes total downloads by week for each format, and goes all the way back to June of 2008.

Update: In response to the comment thread, I’ve added a second version of the graph, overlaid with the running average total downloads. I’m not able to share actual download numbers at this time, but I hope this makes the original more interesting/useful.

Screen shot 2010-10-04 at 10.57.42 AM.png

tags: , , ,

Comments: 26

  1. What’s the big PDF spike around 201020?

  2. @kd — We ran a big promotion (any ebook for $9.99) and it generated a huge response, including a large number of people who hadn’t previously bought our ebooks, so were perhaps most familiar with PDF.

  3. Can you replace the % on the left with real download numbers? This is confusing the way you have this set up.


  4. @Jim — this graph is intended to show relative downloads as percentages of the total, rather than absolute numbers. There is significant fluctuation in total downloads (related to various time-limited promotions) that makes the version scaled to actual downloads much less readable. We also don’t disclose actual numbers of downloads or sales, though we’re already running well ahead of last year’s full-year total with a quarter still to go.

  5. Frank Ch. Eigler

    As other commenters suggest, having the percentage-of-total values decrease says nothing about a “shift to other formats”, as that term suggests a zero-sum transition.

  6. @Frank — I understand what you’re saying, but I disagree with your analysis. Among all files downloaded, PDF is now close to half of the total, whereas when we first began tracking the data, it was closer to 97%. To me the data shows that when we make additional formats available, people increasingly choose those formats instead of (or in addition to) PDF, indicating an increase in demand for and usage of those re-flowable formats. Consumption is indeed shifting — from primarily PDF toward other formats instead or in addition to PDF.

  7. “instead or in addition”

    The data in your chart does not allow us to determine which of these is occurring. (If you excuse the politics, maybe this is like “jobs created or saved” 🙂 But perhaps in this case it doesn’t matter: there’s clearly some demand for re-flowable formats.

  8. I generally download both the epub and the pdf versions since these are universal formats that will work with any reader. It would have been interesting to see how many people download just one version or more than one.

  9. Question: Do you know what caused the big spike in Kindle and Epub downloads in early 2009?

  10. @Nate there is no Epub spike. The Epub curve just get pushed upwards by the kindle spike. And the kindle spike was probably caused by the release of the Kindle DX in 2009.

  11. I always download in pdf, epub, and apk, for possible future convenience (or, in the case of pdf, in case there’s anything I need to print or there are any diagrams that aren’t clear in my reader in other formats). I almost never *read* in anything other than epub, though.

  12. @Andrew Savikas: I have to agree with Frank and Jim here. Percentages are inter-affecting each data, and while I understand your intention to demonstrate market share, it is also important to provide the real number of downloads for growth analysis. I for one am really interested to see the numbers, rather than ratios.

  13. I’ve added another version, with the running average of total downloads overlaid. HTH.

  14. Do your PDFs reflow?

  15. Burt's Bee Spunk

    Can you compare to Kindle formatted downloads that you sell through Amazon?

  16. The “any ebook for $9.99” promotion generated the huge spike as you noted–what percentage of your ebook sales are from the “ebook deal of the day”? Does the recent flattening of the totals curve correspond to the price increase for the deal of the day (from $9.99 to $14.99)?

  17. whoa! the “deal” of the day for an e-book is now _$14.99_?

    every single time i saw it listed at $9.99, i said to myself
    “$9.99 is _not_ a _sale_ price for an e-book, you idiots…”

    but now it’s _$14.99_?

    my word.

    so, well, let me put it here, out loud, at least this one time:

    “$14.99 is _not_ a _sale_ price for an e-book, you idiots…”

    yeesh. what idiots…


  18. Do the numbers include only direct sales from O’Reilly or do they include sales through other vendors as well (for instance from Amazon or Barnes and Noble)?

  19. I also always download all available formats, even though I almost only use the PDF or ePub files. Odd habit, perhaps, but it guarantees my book is always accessible.

    Also, as already pointed out, I suspect you’ll see less eBook of the day sales at the new $14.99 price … I’ve bought several $9.99 books on impulse over the past year, but $15 takes it over that limit. Also, comically enough, some O’Reilly eBooks are cheaper on Kindle directly (e.g. Confessions of a Public Speaker).

  20. I would agree that the $9.99 price definitely made me more prone to an impulse purchase. I’ve been purchasing at least 3 books a month as soon as I heard of the DotD. At $14.99, I’ve been more discerning about what subjects to purchase (have purchased zero at that price). On the other hand, other publishers (packt specifically) don’t offer such deep discounts and don’t offer the books in as many formats. Luckily for consumers O’Reilly and other publishers like packt are shipping sans DRM.

  21. As mentioned a couple of times above, I was quite shocked when I saw the DotD price increase from $9.99 to $14.99. That’s a 50% increase!

    I love the DotD and I bought quite a few books at $9.99 through the DotD before (including many books that were not absolutely necessary for me but just nice to have at $9.99). I don’t think I’ll do that at $14.99 though. I bought zero books at $14.99 so far and I will be very selective with what I will buy in the future.

    I don’t think I’m alone in this so I strongly recommend O’Reilly to reconsider this price increase.

  22. I allow people to name their own price for the eBook version of my book (about running, not techy stuff). The paper version is $12.99.

    I figure that if people can pay what they think a book is worth, they’re more likely to buy. A sale for a little money is better than no sale at all (as long as you pay more than the 50 cents it cost to process your order).

    I suggest $5 is a fair price for the eBook, and I believe most people will be fair. The ones who aren’t fair probably aren’t going to pay anyhow, since they can probably figure out a way to get the book for free no matter what I do.


  23. now we have a “deal of the day” weighing in at _$19.99_… oy!

    > http://twitter.com/timoreilly/status/27104115347


  24. and now the so-called “deal” of the day is up to $21.99…

    > http://twitter.com/timoreilly/status/28057767583

    so it’s become clear they’re running an inelasticity test
    to see how much the “deal of the day” phrase will pay.

    don’t be a sucker. they need to be sent a clear message.


  25. Very interesting, but the images are missing…. any chance of restoring them?