How Many Kindles Have Really Been Sold?

Michael MaceBy Michael Mace, Principal, Rubicon Consulting

Although a lot of people are excited about ebooks, it’s very difficult to get hard information on how the market for them is growing. We don’t even know how many Kindles Amazon has sold, let alone more detailed specifics on the market.

So I was very happy Wednesday when the Book Industry Study Group (BISG) gave details from its recent survey of ebook adoption in the US. The survey was first revealed in January [ed: at the 2010 Digital Book World Conference], but the press release was very sketchy and sometimes confusing. In its presentation at the Tools of Change conference, the BISG gave much more details on the results.

[NOTE: DBW Members can listen to the full presentation from the January release at Digital Book World here.]

My highlights from the presentation:

Ebook usage is growing fast, but it’s still small.

Roughly 2% of American book buyers over age 13 are active ebook users, meaning they obtained an ebook or a reader device in the last year. About half of those were first-time ebook buyers, so the usage of ebooks has probably roughly doubled in the last year. BISG is doing multiple waves in the survey, and says it found a 25% increase in ebook usage just over the holiday season, so it was a pretty good Christmas (and Hanukkah) for ebooks.

The most-used device for reading an ebook is a personal computer (47%); Amazon Kindle is number two (32%), followed by Apple’s iPhone and iPod Touch (21%).

Either there’s something wrong with the numbers, or Amazon hasn’t sold quite as many Kindles as some people think.

What does it mean?

PC leadership is no surprise. There are so many PCs in the US that even a small percentage of PC users reading ebooks will swamp everything else. BISG said that the PC share of ebook reading is declining as other devices grow, also what I would have expected. Apple is closer to Kindle than you might expect, but…

A tidbit that jumped out at me was Apple’s share of ebook usage. Kindle has gotten all the attention, but Apple has about 2/3 the share of Amazon in ebook usage without even trying.

However, before we set off another round of “Apple uber alles” on the web, there are several big caveats:

BISG didn’t report on the number of books bought per platform. Based on my experience at Palm (which had an active e-reading community), I suspect that a lot of those iPhone book readers are pretty casual, buying a few books or publications to kill time when they are bored. I believe Kindle users are probably much more active readers.

(For comparison, about 4% of the Palm OS users in the US were reading ebooks at least occasionally in 2002. That total rose to about 8-10% if you included the Bible — it was by far the most popular ebook. That amounts to about 1.5-2 million ebook users on Palm OS alone.)

Apple and Kindle are also different demographically. After the presentation, BISG told me that Kindle readers are older and more likely to be female compared to Apple readers. What we may be seeing is that if someone already carries an iPhone or iPod Touch, they’re less likely to invest in yet another device just to read on it. Or maybe younger people just find it easier to read on a tiny screen? Either way, I think it’s pleasant that Apple and Kindle are reaching somewhat different audiences rather than just stepping on each other.

The iPhone/iPod Touch installed base is a lot bigger than Kindle’s. So as is the case with PCs, even relatively low ebook usage on the iPhone will add up to a lot of users.

How many Kindles are really in use?

As far as I can tell, Amazon hasn’t released any Kindle device sales figures, other than a quote referring to “millions” of users. Several analysts have jumped on the use of the plural as evidence that at least two million Kindles have been sold. But I think the BISG survey doesn’t support that.

Here’s my math:

–About 2% of book buyers have ebooks and/or ebook devices.

–About a third of them have Kindles (that’s 0.67% of active book buyers).

If 0.67% of book buyers in the US is two million people, then there are 300 million active book buyers in the US. That is the entire US population, including infants and people who don’t like books. I don’t know what the base of active book buyers is in the US, but my guess is it’s not over 200 million, meaning the installed base of Kindles would be about 1.3 million.

It’s tricky to play with survey results when the percentages are this small — the margins of error become very significant. But for now I think the BISG survey raises some questions, and I’m not willing to accept the two million figure for the Kindle installed base without some more rigorous evidence to support it.

[This article was originally published on the Mobile Opportunity blog and has been reprinted here with the permission of Mr. Mace. Click through for more tidbits and analysis.]

Michael Mace is a 20-year veteran of some of the tech industry’s toughest battles, and widely regarded as a strategist who also does tactics, translating big ideas into practical implementation plans. He is a Principal at Rubicon Consulting.

14 thoughts on “How Many Kindles Have Really Been Sold?

  1. Pingback: uberVU - social comments

  2. Michael Cader

    Interesting piece, Michael, but your hunch on number of Kindles sold is demonstrably wrong. For starters, Amazon would be in trouble for saying “millions” in an official filing if the total is less than 2 million. But more particularly, beyond survey data, we have a more important number from Amazon’s last 10-k. They reported “approximately $508 million of deferred revenue” from Kindle shipments in 2009 alone to be booked in 2010 and 2011. The math gets a little tricky, but they sold a minimum of around 2 million Kindle units in 2009 *alone*. All total installed base guesses are just that–guesses; but the idea of a minimum installed device base of 2.5 to 3 million units is quite reasonably supported.

    1. Michael Mace

      Thanks, Michael.

      If the reversal is all Kindles (something that Amazon doesn’t directly say, but does imply), that would point to sales of about two million Kindles before 2010, depending on how steep the sales ramp for Kindle has been. Amazon has a history of doing things to obfuscate revenue from its emerging businesses (Amazon Web Services, for example). But you’re right that this is some separate evidence for sales two million or more Kindles.

      Which raises another question — why didn’t they show up in the BISG data? One explanation would be if a lot of people have bought Kindles (or received them as gifts) but have stopped using them. It’s hard to believe that would have happened with such a new technology, but I think something weird is going on with the numbers and it deserves more investigation.

      1. gous

        The numbers are probably just wrong. There are not 8% of Nook readers for one so that is already a rather serious flaw in their statistics, according to your expanded discussion of their results at We need to be able to compare them with other studies to get an idea if this is just an outlier. Did they reveal anything about their methodology ?

    2. Jack McKeown

      The numbers from our Verso Survey of Book-Buying Behavior (htttp://
      would support Michael Cader’s range: a total of approximately 3.5-4.0 million e-readers in circulation, with 2.3-2.7 million being Kindles.

  3. Pingback: O fim do livro como o conhecemos? « Abrindo o Livro

  4. Barbara

    This is how I interpret the numbers.

    It refers to USERS, RATHER THAN UNITS sold. They may
    have a formula as an ex.there are 1.5 users
    per unit. Newspapers, magazines and books
    are believed to be read by more than one person
    for statistical and adv sales.
    Again–they never use the term–units sold.
    That is why they can make statements without telling lies.

    Personally I think it is silly that they do not report
    actual units–I don’t think it hurts them.
    Recently B&N has also become coy regarding numbers.
    It is annoying.

  5. D

    Also, you are looking at devices only, but then consider that B+N, Kindle and a few others have ebook reading applications on PCs and iPhones…so when a survey user answers iPhone, what platform are they referring to? Some Kindle owners have both a Kindle device and read on their iPhone, and some iPhone users are only using Kindle app to read, does that count towards iPhone’s market share or Kindles? iPhone buyers probably aren’t buying an iphone to read ebooks, thats just a benefit and does not rule them out of buying or owning a purpose built device.

  6. Michael Mace

    Good comments, folks.

    I think the bottom line is that Amazon (and the other ebook players) should share some hard numbers with us if they want companies to invest in their platforms. Even Apple, which is famous for its information control, occasionally gives out tidbits on iPod and iPhone sales.

    Jack, could you give a little more details on the methodology for the Verso survey? How were respondents recruited?

    Also, what was the penetration of ebook readers among the 62 million people you identified as avid readers? I think that would be the audience of greatest concern to publishers who are worried about cannibalization…

    1. Jack McKeown


      The methodology is shorthanded in the slide show which is available online @ (also via links on the DBW and ABA websites), which I assume you have seen.

      Essentially, our surveys piggyback on the Verso Digital vertical ad netowrk of 110 million monthly uniques. The sample pools are large, over 6,000 respondents, statistically weighted to mirror the U.S. 18+ population. It yielded a relatively low margin of error at 95% probability threshhold.

      The relevant findings for extrapolating e-reader numbers are in the last section of the presentation. One slide in particular polls active book buyers and determines that 2.9% currently own e-reader devices (as of 12/09). That translates into 135 million 18+ book buyers x .029 = 3.9+ million total e-reader owners. The Kiindle numbers are extrapolated from that total number, an admittedly less-than-rigorous method, but all we have given the lack of hard data from Amazon. The e-reader results were based on polling all book-buyers, not just the avids. Your last question might present a worthy one for future surveys, particularly in measuring the dynamic changes in e-reader ownership/demographics over time.

      1. Barbara

        Has anyone looked at this from the capacity of their manufacturer
        or component suppliers? You would think they might want to brag about
        these great new products, but everyone is quiet.

        Also the Nook ran late and the iPad is running late–are there supply issues,
        manufacturing issues, quality issues, or did everyone just think they could be
        be turned around quicker. Again–no one is talking and journalists don’t seem
        very curious–tend to take what they are told in this field.

        1. Michael Mace

          Hi, Barbara.

          Well, Apple shoots any supplier who talks. Maybe Amazon and B&N have adopted similar policies.

          Regarding the late schedules, almost every new-category product that I’ve ever been involved in has slipped its schedule. There are so many unknowns that it’s almost impossible to predict an accurate schedule. The wise companies build that into their plans, and also avoid promising an exact shipment date until they actually have the product in hand.

  7. Pingback: Dear Author: Digital Books from the Consumer’s POV | Digital Book World

  8. Pingback: Los vericuetos de la P de Publishing. Deshaciéndonos del autor por la razones equivocadas — Filosofitis


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