State of the Computer Book Market 2008, Part 3: The Publishers

In this third installment, (see part one and part two; part four to come later this week), we will look at how Publishers fared in 2007 when compared to 2006. The chart below shows our dashboard view of the Large publishers’ results for 2007. The most notable factor is that Wiley continues to hold the leading spot as the largest publisher, with 30% market share of units sold, while Pearson lost 2% market share and O’Reilly gains 1%. (We’ll look at revenue share later in the analysis.)

2008 Pub Share 2007 Pub Share
Leading Computer Book Publishers for 2008 Leading Computer Book Publishers for 2007

You may not recognize the names of all the top publishers, because they are actually conglomerates of many smaller publishing imprints that they’ve acquired or created over the years. The imprints are the familiar consumer-facing brands. For instance, when you purchase a book from Peachpit or Sams, you typically see Peachpit or Sams on the spine, not Pearson, even though Pearson owns both companies. One stand-out change is that Thomson and its computer-book-oriented imprints are now called Cengage. Also, all the imprints that make up O’Reilly “publisher view” are part of a distribution partnership and are not owned, in any way, by O’Reilly.

Let’s look at the top Publishers and how they performed year over year. The following table provides some interesting comparative data.

Publisher ‘08 Units ‘07 Units ‘08 Title Count ‘07 Title Count ‘08 Efficiency ‘07 Efficiency
Wiley 2,112,165 2,140,601 1,434 1,415 1.54 1.47
Pearson 1,813,401 2,014,514 1,934 1,974 0.98 0.99
O’Reilly 1,163,965 1,143,279 846 818 1.44 1.36
Microsoft 711,244 791,340 357 327 2.08 2.35
McGraw Hill 313,970 358,699 430 404 0.76 0.86
Apress 228,344 235,591 396 336 0.60 0.68
Cengage 167,472 212,680 580 610 0.30 0.34
Reed Elsevier 167,480 157,157 366 356 0.48 0.43
Sum/Avg 6,678,041 7,053,861 6,144 6,240 1.02 1.06

So what is notable from this data? First that these top 8 publishers are down – 375,820k units from 2007 to 2008. Only O’Reilly and Reed Elsevier saw modest gains in 2008. Seven out of the eight top publishers had more titles making the top 3000 list in 2008. Wiley, O’Reilly, and Reed Elsevier saw their efficiency improve in 2008 while the other large publishers saw their efficiency decrease.

A note on Market Share versus Title Efficiency

A typical indicator of publisher performance is market share of units sold, which is what we’ve been looking at so far. Perhaps a better measure is how many titles published it takes to get a comparable share of unit sales. This is the ratio of title share to unit market share. Think about it this way: if a publisher has 15% of the titles appearing in the Bookscan Top 3000, and gets a 15% share of units sold, they will have a ratio of 1:1, expressed as a title efficiency of 1.0. A publisher with 20% of the title share, and 10% of the unit share would have a .5 efficiency. An efficiency of 1 is the market average: 100% of the title count delivering 100% of the unit sales. A publisher that achieves its share with fewer titles will have a higher ratio. Publishers under the 1.0 threshold typically have many titles in the Bookscan data, but they are not selling many units. A note of caution though. Some publishers have many evergreen titles [mentioned in the last post] that can skew this data. Typically older titles sell fewer unites each subsequent year. But this is not always true, and some titles continue to sell like they are newly released. (Head First Design Patterns is one example), And can still sell more than some brand-new titles. So efficiency could be thought of as a “frequency ratio” rather than efficiency – because it is very efficient to publish a title and have it sell for years and years.

A note on Evergreen Status

In the table below has a view of Imprints that have a percentage of evergreen titles. So here is how we can up with evergreen status. We assigned points to titles that had copyright dates older than 16+ years [most points], 10-16 years, 5-10 years and less than 5 years. After assigning points for each title, we were able to see what percentage of evergreen titles each imprint had in the top 3000 for 2008. It’s interesting to note that the top three evergreen imprints have a strong academic heritage.

Imprint % Evergreen
John Wiley 12.66%
Addison Wesley 12.23%
Prentice Hall 11.48%
O’Reilly 9.57%
For Dummies 9.03%
Microsoft 7.81%
Sams 7.58%
Que 7.52%
McGraw-Hill/Osborne 7.43%
Peachpit Press 6.93%
APress 5.25%
Wrox 2.51%

Now that you have a basic understanding of title efficiency and evergreen status, let’s look further into the 2009 results for the imprints and drill in on the top three publishers: Wiley, Pearson, and O’Reilly. Click on any chart to get a bigger image. The reason this is important, is that you typically see the imprint name on a book when you purchase it. (You’ll likely see the publisher inside the book on the copyright page, except for O’Reilly because our other imprints are distribution partners. That is, O’Reilly provides some sales and distribution services to these partners, but they are not owned or home-grown parts of their business, which is the case for Pearson and Wiley.)

Wiley Pearson
Computer Book Sales 2008 - Wiley Publishing Computer Book Sales 2008 - Pearson Technology Group
O’Reilly Media
Computer Book Sales 2008 - O'Reilly Media

Now that you have an idea of the imprints that make up the largest three publishers, let’s throw all the imprints together and then look at their respective market share. The following chart shows the top 10 “imprints” and how they stack up against each other. From this imprint-view, you’ll notice that O’Reilly has the second largest market share behind “Dummies” and Microsoft Press.

Top 10 11-20
Computer Book Sales 2008 - Top 10 Imprints Computer Book Sales 2008 - 11-20 Top Imprints

So what do the graphs tell us? The first notable thing is that O’Reilly moved ahead of the pack as the second largest imprint. The percentage in the pie chart is just for the titles represented in the chart, not market share. In the second ten, Wiley’s Sybex imprint moved ahead of Prentice Hall for the third spot in the list.

Before analyzing imprints by category, let’s revisit the data with Dollars rather than units. We have a fairly easy way of calculating this: units sold * listprice = dollars. Granted there are discounts, promotions, and other things that affect the precision of this, but it is pretty close, if nothing else you can think of this as “retail value.” So here are the top imprints from a revenue perspective. Again, this is at the imprint level and from a dollar perspective. As you can see compared to the publisher share above the leaderboard quickly changes. O’Reilly becomes the number one revenue-producing imprint, followed by Microsoft Press and then Dummies. Que disappears from the top ten and Prentice Hall climbs into the list.

Computer Book Sales 2008 -- Top Imprints - Dollars

Imprint Analysis by Category

Now that have seen high level picture of what imprints did in 2008, let’s take a look at which categories each of them publishes in and where their strengths lie. Dummies and O’Reilly appear to have the most diverse publishing programs as they are not at the bottom in any category. Dummies is clearly the leader in Business Apps and Consumer Operating Systems, while O’Reilly and Microsoft are neck and neck in the Systems and Programming category. O’Reilly enjoys a market position ahead of Microsoft Press mostly because of its strength in the Web Design/Development and Digital Media areas.

Publishers’ Category Strength
Computer Book Sales 2008 - Imprint Strengths

Categories and the Publishers who dominate them

The following category images are for 2008, and the tables have each publishers’ count of titles and sum of units. The top titles are also listed for each area in 2008.

Category: Systems and Programming

In this category you can see that Pearson has the largest market share among the publishers with O’Reilly second. If we drill into the imprint level, the picture changes a bit. The top six imprints are O’Reilly at 16.08%, Microsoft Press at 14.05%, Addison-Wesley at 8.42%, For Dummies at 7.41%, Sams at 6.25%, and Apress at 5.85%.

Computer Book Sales 2008 - Publishers - Systems and Programming

As you can see in the table below, Pearson’s and Wiley’s market share is primarily driven by title count, whereas Microsoft Press sells roughly twice as many units per title — they are the most efficient publisher in this space. O’Reilly’s performance is a mix of both. That is, we have quite a few titles, but our efficiency is also significantly above the market average.

Sys & Prog – Publisher Market Share ( 01/01/2008 – 12/31/2008 )

Publisher Units Title Count Units/Title Efficiency
Pearson 587,449 967 607 0.94
O’Reilly 452,952 455 995 1.54
Wiley 407,243 539 756 1.17
Microsoft 287,598 205 1,403 2.17
McGraw Hill 123,291 213 579 0.9
Apress 95,845 187 513 0.79
Cengage 51,162 200 256 0.40
Reed Elsevier 43,021 189 228 0.35

Note: This category family contains “programming languages” where more units were sold in 2008 than in 2007. The leading titles and publishers for Systems and Programming in 2008 were:

  1. RMC’s PMP Exam Prep: Rita’s Course in a Book for Passing the PMP Exam
  2. Addison Wesley’s Cocoa
  3. Wiley’s CCNA: Cisco Certified Network Associate Study Guide
  4. McGraw Hill’s: CISSP Certification All-in-One Exam Guide, 4th Ed.
  5. Cisco Press’s CCNA Official Exam Certification Library
  6. O’Reilly’s Learning Python

Category: Web Design and Development

In this category you can see that Pearson has the largest market share among the publishers with O’Reilly second. If we drill into the imprint level, the picture changes a bit. The top six imprints are O’Reilly at 19.17%, Dummies at 12.29%, Peachpit Press at 10.26%, Sams at 5.83%, Friends of Ed at 5.42% and Apress at 4.92%.

Computer Book Sales 2008- Publisher Share - WebDesDev

As you can see in the table below, Pearson has the most titles and their performance is strong in this category. In this category, most of the top publishers are above the title efficiency average of 1.0. (This suggests that there are a lot of second-tier publishers with lower efficiency who don’t show up in the table.)

Web Des & Dev – Publisher Market Share ( 01/01/2008 – 12/31/2008 )

Publisher Units Title Count Units/Title Efficiency
Pearson 359,992 275 1,587 1.20
O’Reilly 294,838 183 1,611 1.48
Wiley 279,128 205 1,362 1.25
Apress 99,298 146 680 0.62
Microsoft 37,258 31 1,202 1.10

The leading Titles and Publishers for Web Design and Development are:

  1. Peachpit’s HTML, XHTML, and CSS: Visual Quickstart
  2. O’Reilly’s Dreamweaver CS3: The Missing Manual
  3. Wiley’s Building a Web Site For Dummies
  4. New Riders’ Don’t Make Me Think: A Common Sense Approach to Web Usability
  5. O’Reilly’s Head First HTML with CSS & XHTML
  6. Wiley’s WordPress For Dummies

Category: Business Applications

In this category you can see that Wiley has the largest market share among the publishers with Microsoft Press has moved ahead of Pearson for second. If we drill into the imprint level, the picture changes a bit. The top six imprints are Dummies at 28.21%, Microsoft Press at 18.37%, McGraw Hill/Osborne at 6.78%, O’Reilly at 6.19%, John Wiley at 5.76%, and Peachpit Press at 4.89%.

Computer Book Sales 2008 - BusApps

Bus Apps – Publisher Market Share ( 01/01/2008 – 12/31/2008 )

Publisher Units Title Count Units/Title Efficiency
Wiley 686,329 351 1,955 1.73
Microsoft 286,495 97 2,954 2.62
Pearson 259,323 298 870 0.77
McGraw Hill 114,839 103 1,115 0.99
O’Reilly 102,034 75 1,360 1.2
Cengage 43,847 177 248 0.22
Apress 20,409 41 498 0.44

The leading Titles and Publishers for Business Applications are:

  1. Wiley’s Microsoft Office 2007 All-in-One Desk Reference for Dummies
  2. McGraw Hill’s QuickBooks 2007 The Official Guide
  3. Wiley’s Excel 2007 for Dummies
  4. Microsoft’s Microsoft Office Excel 2003 Step by Step
  5. Wiley’s eBay For Dummies
  6. Wiley’s Excel 2003 for Dummies

Category: Consumer Operating Systems

In this category you can see that Wiley has the largest market share among the publishers with Pearson in second at 13%; O’Reilly has moved ahead of Microsoft Press for the third spot. (Apologies: the Pearson % is cut off in the graph.) If we drill into the imprint level, the picture changes a bit. The top six imprints are , For Dummies at 31.99%, O’Reilly at 15.47%, Microsoft Press at 9.52%, Que at 7.83%, Peachpit Press at 6.70%, and Wiley’s Visual at 6.21%.

Computer Book Sales 2008 - ConsOpsys

As you can see in the table below, Wiley has the most titles and a relatively good efficiency rating. Microsoft Press also has a very healthy title efficiency rate. O’Reilly is the only other publisher with an efficiency rating higher than the market average.

Cons Opsys & Dev – Publisher Market Share ( 01/01/2008 – 12/31/2008 )

Publisher Units Title Count Units/Title Efficiency
Wiley 500,979 150 3,340 1.46
Pearson 201,347 114 1,766 0.77
O’Reilly 177,625 41 4,332 1.9
Microsoft 98,752 21 4,702 2.06
McGraw Hill 52,308 52 1,006 0.44
Cengage 34,118 53 644 0.28

The leading Titles and Publishers for Consumer Operating Systems are:

  1. O’Reilly’s Mac OS X Leopard: The Missing Manual
  2. Wiley’s Windows Vista For Dummies
  3. Microsoft’s Windows VIsta
  4. Wiley’s Windows Vista For Dummies: Special DVD Bundle
  5. Wiley’s Windows Vista All-in-One Desk Reference For Dummies
  6. WIley’s Windows XP For Dummies

Category: Digital Media

In this category you can see that Pearson has regained the top spot with Wiley falling to second and O’Reilly moving up into the third position for all publishers in the Digital Media category. (Apologies: the Wiley %[23%] is cut off in the graph.) If we drill into the imprint level, the picture changes a bit. The top six imprints are Peachpit Press at 24.50%, For Dummies at 12.94%, O’Reilly at 12.65%, Focal Press at 11.27%, and Adobe Press at 9.88%, and New Riders at 7.79%.

Computer Book Sales 2008 - DigMedia

As you can see in the table below, Pearson has the most titles and a relatively good efficiency rating. O’Reilly has the highest efficiency rate and best average units per title.

Digital Media – Publisher Market Share ( 01/01/2008 – 12/31/2008 )

Publisher Units Title Count Units/Title Efficiency
Pearson 379,077 205 1,849 1.42
Wiley 194,632 150 1,298 1.00
O’Reilly 113,702 49 2,320 1.78
Reed Elsevier 90,586 98 924 0.71
Cengage 19,872 79 252 0.19

The leading Titles and Publishers for Digital Media are:

  1. Adobe Press’s Adobe Photoshop CS3 Classroom in a Book
  2. New Rider’s The Adobe Photoshop CS3 Book for Digital Photographers
  3. WIley’s Photoshop Elements 6 For Dummies
  4. New Rider’s The Photoshop Elements 5 Book for Digital Photographers
  5. O’Reilly’s Photoshop Elements 6: The Missing Manual
  6. Peachpit’s Photoshop CS3 for Windows and Macintosh

Next Up, The Languages


Get the O’Reilly Programming Newsletter

Weekly insight from industry insiders. Plus exclusive content and offers.

  • Ugh. You construct a self-reinforcing but arbitrary and technically bogus hermeneutic when you insist on breaking things down by programming language. I dread the next installment, especially from an influential company that once arbitrarily declared “no lisp books!”. Language should not be so central to the analysis of what is important and where language figures into the analysis, you guys have a history of talking about it in inane, disruptive ways (where your pronouncements then have substantial economic influence on the practice of software engineering).

    I do understand that as a variable for deciding what manuscripts to shop for, where and how to market, etc. that programming language is so significant as to not be ignored. But there must be more thoughtful and responsible ways to do it if we decline to neglect paying attention to how your market analysis and manuscript/marketing choices influence practice.


  • Thomas-

    As I try to make very clear. This analysis is about Book Sales. We are not trying to rate the quality of any technology or language, nor are we trying to say which is best for building any sort of software or system.

    I am sorry you feel angst about the next post. Are you sure we have “substantial economic influence on the practice of software engineering” when some of the languages that were way down on our list have moved up considerably. And some that were moving up in unit sales have moved back down. I actually try my hardest to just report on the news of books sales in a market. And there are a fair share of programming language books.

  • Mike

    So, in “what it looks like from here…” mode. I don’t claim to have perfect knowledge of how your impact plays out:

    You’re talking about book sales and, as I tried to indicate, I fully recognize the importance of programming language as a variable there. For example, if you want to have some influence on the shelf inventory of retailers the fact is that people go to those stores with the thought in mind “I need an intro language X” (and such) and in that sense your analysis is likely very important to the firm. And that kind of thing – I don’t just mean shelf inventory but even, say, allocation of ad budget.

    The problems I’m grousing about arise because O’Reilly is more than just a bookseller. It’s also a conference host and a “pundit” blogger operation. And, although it’s fuzzy – kind of in the society of capitalists rather than in capitalism per se – there’ also just Tim in the social milieu of the vcs, and the talking heads, and the highly paid free-agents that travel among them and all that.

    So, the net effect, looking at the big picture, is that you get little constellations like radar encouraging programmers to dedicate their ongoing learning efforts to say, Ruby (on or off the rails), and in conjunction you get conference session hours thrown over to that momentum, and – one interpolates – you get some buzz about Ruby (or whatever example we like – I’ve nothing particularly against Ruby) among the signers of big checks. And what this means is that – in practice (that is, in practical effect) – Tim is performing a senior lead engineer role on a very massive scale project.

    Only – wait a minute – is that in the slightest bit rational? I mean: of course it isn’t. It shouldn’t be that way, yet it is.

    Now, it’s not fair in my understanding of things to O’Reilly & Assoc or to Tim the guy to be in this bind. He and the firm are, after all, running a few straightforward basic games in the markets: publisher, pundit, conference host, etc. It’s a quasi-“sum of games” in the game theory sense. Wouldn’t it be nice and wouldn’t we wish for a world in which a business could engage in various separate activities like that and have it just be a simple sum of orthogonal issues. But – that’s just not realistic.

    My judgment, reading slashdot over the years, watching the vcs, watching O’Reilly, etc. – is that the games aren’t a true sum but that there’s a lot of entanglement. Put more simply: when O’Reilly says “books about language X are uptrending” *in the way O’Reilly talks about that fact* the result is a non-linear correlated change in investment practices on a very large scale and range of scales – everything from what individual geeks decide to learn in their spare time to how major VC funds pick and choose or big think-tanks sum up the market. And it’s no accident that O’Reilly is in that position as evidenced by things like the acquisition of “release 2.0”.

    And so the firm is firmly entrenched in *doing* large-scale software engineering and when you talk very well about book stats as you’re about to, you are very much a component in O’Reilly’s software engineering activities.

    And, from my hacker, alpha-geek(-wannabe?) perspective? O’Reilly gets the engineering questions – the ones that matter in terms of the social responsibilities of software engineering – dead on Wrong way too much of the time.

    The old “lisp grudge” (the ‘no books on lisp’ whine) is just a small and hopefully familiar enough example. For whatever amusing reason such declarations were made the fact is they had the impact of a kind of censorship and a kind of rabble rousing in service of an irrational prejudice, the net effect of which was to further degrade the practice of software engineering in our society. And this, allegedly, in service of a firm in the business of spreading useful information.

    So, yeah. I dread your next installment even while I have great respect for your technical skills. It’s “here we go again” on a historically quite destructive social dynamic.


  • I probably bought about 100 computer books this year, and would love to own like literally every one if I ever can some day, so I found this report to be amazingly insightful and useful.


  • Jim Gleaves


    I’ll lead with a shameless plug. If you bought 100 books this year but want to own all of them then you should seriously consider subscribing to Safari Books Online. I work for Safari, and we like your kind of person.


    Have you considered cross-referencing any of this information on book purchases with information in Safari? I think that the online availability of information is really changing the publishing game in some surprising ways.

    Consider this–paper books compete with each other for sales based on things like price, copyright year, etc. However, in the online version different criteria come into play when deciding what books to read. In Safari issues like searchabilty and clarity of writing are much more significant. I suspect there is a real difference in which books are most popular in paper form vs. digital form, but don’t have any data to back this up yet. I’d be happy to help in running any usage reports that might be useful.

    To the best of my knowledge we don’t have any authors yet who seem to be taking this into account and writing for the online version. But as the online version becomes more standard smart authors will change with the times.

    You can contact me at Safari tech support if you’re interested.

    Jim Gleaves

  • for proj