The Ultimate Effect Of Google E-Books: A New Ad-Supported Model For Books


Credit: Google ebooks

Brace for impact. What I’m about to say is going to make a lot of people angry, including some of the people at Google (NSDQ: GOOG) who are going to rightly point out that when they pre-briefed me on today’s Google e-books announcement, we never once discussed ad-supported reading.

Instead, they told me all about their plan to establish a set of tools that will offer e-books to people looking for book information through Google’s search engine. They explained that this will make it possible for the millions of people who conduct book-related searches every day to have easy access to 3 million books — some out of copyright, some out of print but under copyright, and a full range of in-print titles including bestsellers. They also described how independent booksellers will be able to use the same set of web-based commerce and reading tools to build their own branded e-book stores to finally extend their brick-and-mortar customer relationships into the digital space.

Since then, I’ve spoken to half a dozen reporters who were also pre-briefed and they have all had a similar set of questions: can Google compete against Amazon? (NSDQ: AMZN) (No, but it can compete against Barnes & Noble.) Is it too late to make a dent in a mature market? (No, less than 10% of online adults in the U.S. read e-books; there’s plenty of room to grow.) Is Google’s cloud-based strategy unique? (Yes and no: It supports all devices except the Kindle, but the Kindle platform actually supports as many devices as Google will.)

All of that is true, and I’m genuinely glad to have Google enter this market because it will be reaching potential customers at a unique point in their book-buying journey: at the point of web search, not at the point of searching the bookstore. This means many things you didn’t realize a book can help you with — overcoming depression, remodeling a bathroom, making friends and influencing people — will now be surfaced alongside all the YouTube and other results Google will offer. This is a net plus for books.

But the ultimate effect of Google e-books, if Google knows what’s good for it, will be the creation of an ad-supported publishing model. This is where the anger starts. Because I’m putting words in Google’s mouth — words they may or may not be thinking but certainly words they don’t want to say out loud — but mostly because I’m proposing something that is so counter to the philosophy and mindset of the publishing industry that I may be shunned for even suggesting it.

Opponents of my suggestion will conjure a worst-case scenario, where Dickens’s, “It was the best of times, it was the worst of times,” could end up sponsored by a Google AdWords campaign that reads, “Is your day feeling like the worst of times? Try our new sports drink to get your afternoon back on track.” I’ll agree this example may be atrocious, but I can just as easily identify dozens of very smart examples. For example, not only would Kohler and Home Depot want to sponsor your browsing of a book on bathroom remodeling, as a reader you may actually appreciate the link to the relevant plumbing resources.

Before anti-advertising types crawl all over me, let me explain the most important reasons why this makes sense. First, books are the only medium left not significantly sponsored by advertising. From the Android Angry Birds game app to Pandora music streams to to the venerable, advertising is essential to the success of nearly all media — analog and digital. The only reason book advertising has not happened is that the economics of distributing books have required that people pay for them — in a way they have never paid for the newspaper, magazines, or even music, where a majority of listening has always been radio-based. If you make people pay the full price of a book’s creation and distribution, you can hardly expect them to endure advertising. Plus, books last for such a long time that an ad placed 25 years ago in my copy of The Hunt For Red October would be laughably irrelevant today.

That has all changed now. Since Google intends to provide its books from the cloud, it can deliver ads that are timely and targeted. And the economics of publishing are swiftly moving away from an analog production model. (I blogged about this to much ado last month), which means that soon, we will no longer need to force the entire cost of a book on the buyer of the book, but instead can extract value from the reader of the book, in direct proportion to the value they get from it. In other words, the more pages they read (the more value they get), the more ads they see and the more value the publisher and author receive.

Also, consider this: The two most common ways people acquire books today are borrowing from a library and borrowing from a friend. There are several reasons for that; economically, a principal reason for this is that until you’ve read a book you can’t be sure whether it’s worth the price printed on the book’s jacket, so you prefer to get the book for free when possible. But how can you get people to try a book to find out if they like it before buying it?

In the e-book world, Amazon and now Google can let you sample the first chapter or two (or three) for free. And while that’s a good way to get someone interested in a book, advertising is an even better solution. Imagine you find a new book, highly recommended by your Facebook friends or a blogger you trust. Instead of a limited free trial of the first chapters, from which the publisher and author get no recompense if you don’t ultimately buy, what if you were offered an ad-supported read of the book instead? As you read, the meter starts ticking and revenue accrues directly to the publisher and author. If you don’t like the ads, you can buy the book at any time to remove them. But even if you choose to read the whole thing without buying it, it’s an easy thing for Google to figure out how many ads you have to see for it to generate enough revenue to replace the cost of buying the digital book.

I’ll stop there, I have a hundred more justifications for why this is the next logical step for the industry, why Google is perfectly poised to do it. And, I can anticipate all the complaints from publishers as to why they will hold strong and reject this extra (easy) money. But it will happen. And by raising this flag and starting the verbal war that will ensue, I have done my part to ensure we get to the future sooner rather than later.

James McQuivey is an analyst at Forrester Research, where he serves Consumer Product Strategy professionals. James blogs here.

This article originally appeared in Forrester Research.


James McQ

Hal, I appreciate your commitment to your points, but you’re saying I’m defending advertising and pretending it’s not in trouble when I have said no such thing. I am not promising that publishers will suddenly thrive just because they add an advertising model. As I indicated in my responses above, this is just one of many things publishers will have to do — mostly to diversify their financial model because no single revenue source — advertising included — is a silver bullet.

I’m not holding up Hulu, for example, as the salvation of the video industry and suggesting book publishers should do the same. I’m holding up Hulu to refute your claim that people will do everything they can to avoid advertising. That’s not true. They will do everything they can to consume content, including enduring advertising. Is that sufficient to then bet the whole farm on advertising? No.

The only solution is to pursue all solutions simultaneously. And book advertising is a really smart idea for all the reasons I’ve identified above. Not to replace revenue, but to seek it elsewhere, while changing the model for discovering books and allowing them to be shared.


“The easiest one to point out is that 3x as many people watch content on Hulu than do on Netflix in a typical month…”

Huh. So, given that (as previously mentioned) Netflix makes $2.2B a year while Hulu makes $240M a year (per each of their most recent quarterly reports), you’re implying Netflix makes about 27 times as much per viewer than Hulu does.

And you’re advising publishers that this is a good model to follow? “Follow me, boys, I can cut your revenues per reader by 96%!”

“I do disagree on your one hopeful note that “advertisers are wising up” because, unfortunately, they don’t seem to be. At least not most of them.”

That seems a very high threshold. You appear to be implying you won’t concede advertising is in trouble until both revenue and percentage of GDP falls by 50%. By this standard, the Great Depression just wasn’t that bad, since unemployment hit 33%, but hey, “most” people still had a job.

“(C)able cutters don’t currently number more than a few tens of thousands.”

Well, I must personally know a fair proportion of them, then. I suppose that’s possible. I don’t usually think of myself as being in that kind of a bubble, but… {shrug}

Then again, I was among the first 1000 or so users who did computer-based chat ever… on Bitnet’s RELAY in 1986 (numbers per Jeff Kell, the writer of the program). It felt like a big, globally-distributed crowd back then, so subjectivity may go a long way.

James McQ

Seth, the data I cited are not only true, they are well understood by the industry. What you’re not considering is that the vast majority of books read in any given month are not bestsellers. This is in contrast to the movie business. In fact, a large chunk of the books read in any given month were not purchased in that month at all. So when you look at lists of publisher market share, they’re reporting on share of the books sold that month, but not the books read that month, which is a much larger number, thanks to loans and libraries. So you lay out a library scenario where only .1% of the population can access the bestseller in the library while 3x as many people want to read it. That is true, but meanwhile, the library is loaning out dozens more books at the same time, books that aren’t on the bestseller list (even if some of them used to be). So libraries can easily supply more books than bookstores simply by caching so many of them — from bestsellers to classics to random others — and making them all available for free.

The same is also true for book loaning. The books people loan are not always bestsellers or even recent sellers. People loan classic texts from their personal libraries all the time (If I had a quarter for every time I’ve gotten someone interested in Orson Scott Card by loaning out Ender’s Game…). But the real issue with book loans is that you’re assuming 50% refers to the percentage of books read. It doesn’t. This is the most common way people get books that they read. For people who read a lot, it will be a small portion of their reading (me and you), for people who read a little, it can be most or even all of their reading. That’s how it reaches the level of most common way people get books they read, because more people do it than anything else. However, for you and me, when we buy 10x as many books as we borrow, though we are a minority of readers in number, we are disproportionately valuable to the industry.

James McQ

This is valuable dialogue, Hal, and I won’t disagree with most of what you say, especially the challenge associated with establishing the causality of advertising’s success. I do disagree on your one hopeful note that “advertisers are wising up” because, unfortunately, they don’t seem to be. At least not most of them.

As for some of your suppositions about consumers and what they prefer, I do, in fact, have research to show what consumers value and how they behave. The easiest one to point out is that 3x as many people watch content on Hulu than do on Netflix in a typical month, so it can’t be the case that people are going to great lengths to avoid advertising. Instead, people go to great lengths to get top content, including enduring advertising when necessary.

Oh, and by the way, though this is a different discussion, cable cutters don’t currently number more than a few tens of thousands. So that may not be the best comparison to make yet, at least until cutting picks up.


” (T)hank you, Hal, for calling me young.”

You’re welcome.

“I will remind my college son that even though he thinks I’m ancient, there are smart people out there I have fooled, at least on this point.”

Hey, everything’s relative. Given the data I was presented with (someone willing to say, “The only reason book advertising has not happened…”, when book advertising has already happened, and then failed in the market), I thought the more likely, give ’em the benefit of the doubt scenario was that you weren’t old enough to be reading/buying books during the relevant years. If that’s not accurate, well… Everything is provisional, pending better data. {shrug}

“I couldn’t conclusively prove that he was wrong because what was lacking was the measurement necessary to identify the waste. Now that we have more measured media, we are seeing advertising fall because advertisers are more willing and able to cut out the fat.”

On this, I’m afraid I have to disagree.

Here’s the problem, as I see it — there are basically four scenarios. You could draw them as a straight Cartesian grid, but that can’t really be done in text. So, to summarize:

1) +a/+s — increased spending on advertising leads to increased sales
2) +a/-s — spending on advertising increases, but sales fall anyway
3) -a/+s — spending on advertising is cut (or is low), but sales rise
4) -a/-s — spending on advertising is cut (or is low), and sales fall

The Conventional Wisdom — and the implicit transaction sellers of advertising offer — is that 1) and 4) are how the world works. The problem is, 2) and 3) happen often enough to call that into doubt. Political candidates who spend lots but lose (Meg Whitman and the lost $170 million comes immediately to mind). The Blair Witch Project coming out of nowhere and grossing $140 million. Starbucks becoming a global brand from Schultz buying the company in 1987 to 2007 before running their first TV ads. The 12.8 million hits on Google (no exaggeration) for the search term “disappointing sales despite ad campaign.”

I would say advertising is a classic example of partial reinforcement, right up there with lottery tickets. Because the results are so wildly variable, individual examples to fit one’s premise can always be found. But that’s precisely it. If scenario 1) above held true, and if the relationship between sales and ads wasn’t completely random, such examples couldn’t be found.

And the advertisers are wising up.

That’s why the “more measured media” you promote are driving ad buys down. The more data, the more precise the measurement, the more it’s obvious ads don’t deliver what they promise.

As to digital media being less expensive to produce — yes, of course. But digital media also have made it much easier to avoid the ad culture. I don’t think piracy is solely driven by the appeal of “free.” I think it’s also driven by “no commercials.”

To put it another way, I think ad-free Netflix makes $2.2B a year while ad-supported Hulu makes $240M a year for a reason.

People will change their behavior to avoid ads almost as strongly as they will to avoid taxes. One might almost think the person on the street sees them as equivalent.

Again, this is nothing new to anyone who’s been reading that well known anti-advertising rag AdAge during the last five years or so.

When you take the stance, “Declining ad sales are a good thing for the marketing business because it just means we’re getting more efficient!” it rings as falsely as cable companies saying, “‘Cable cutters’? I don’t think they exist.”

Seth Godin

Thanks for answering my query on this, but I confess to being puzzled.

You wrote:

We surveyed them in May/June of this year. The top two ways they acquired the book(s) they read in that time were: 50% a loan from a friend, 38% from a public library (I know, cool, right?), and thanks to rounding, #3 is also 38%, buying a book from a chain bookstore (e.g. Borders, Barnes & Noble).

Here’s my problem (as someone who loves books, buys books and loans them like crazy, also as someone who often has trouble with tricky math questions):

If we assume that a loaned book is loaned out once, that would mean that you would need bookbuyers to be loaning as many books as they are buying. For example, looking at one book-reading cycle, if there are 6 people in a town, for the three borrowers to read, they’d have to borrow from the three buyers, right?

Have you ever met anyone, ever, who loans more books than he buys? I can’t conceive of how that could be true. The only explanation other than a flawed question is that there are hyper-loaned books, books that are loaned out hundreds of times.

As for the library stat, again I’m a bit confused. Let’s assert that many of the books that are read are bestsellers. If a town has one copy of a bestseller for every 1000 residents (that would be 3 copies in a town of 3,000, which feels right, cause it’s not a Blockbuster rental store), that means that in a given month, only .1% of the population can read that book from the library, but in fact the public is buying a million copies of that blockbuster (and probably reading it) which would be .3%, or three times as many.

Sure, there’s a long tail at work, particularly at the library, but it is going to be skewed by the short head.

Clearly, both thought experiments are rife with errors, but as someone who has spent his professional life buying and selling books, neither stat passes the smell test.


Seth, I agree that the existing CPM model won’t generate significant revenues, but once the principle of ad supported book publishing is accepted it opens up a lot of new opportunities; sponsorship, product placement, affiliate sales, etc. For example, how much would the French Tourist Board pay to sponsor a book like ‘A Year in Provence’? Would $0.25 for every free download of the ebook be reasonable? Sounds OK to me, especially compared to AdWords rates. That would translate into $250k for a million downloads rather than $5k. Similarly a book on fly fishing in Scotland could drive affiliate sales to hotels in the area, or maybe companies like Apple would pay for product placement in business books?

These kind of commercial models could fundamentally change book publishing, but digital formats are doing that anyway, so maybe its a good time to start with a clean sheet of paper and consider all revenue options?

David A Graves

James, you’re right. It should be noted that AdSense serves up a combination of CPC and CPM ads, choosing which has the higher return. For some books, particularly business or specialized titles, the CPC rate could be quite high. Seth, I’d expect your effective CPM’s to be really high! If it generates a lead, a book click could be worth a lot. I think books and most content for that matter will evolve into the kind of mixed model like cable TV is today – you pay something to get a bundle of stuff for “free”, you see ads in some and pay more for things without ads. You can still get free over the air TV, which actually looks great in digital, but that only a minor percentage of viewers will ever find. Back to books, there is still a mountain of specialized information that’s locked up in paper and anything that moves that toward online access is welcome.

James McQ

Great comments, I don’t want to miss all the points everyone’s bringing to the table, so bear with me as I wade through the issues. First: thank you, Hal, for calling me young. I will remind my college son that even though he thinks I’m ancient, there are smart people out there I have fooled, at least on this point.

To that point, note that although book advertising has happened (in genre fiction that has rapid turnover so the problem of timeliness is less severe), the fact is that it doesn’t happen today, even in that type of genre precisely because we have more immediate and targetable opportunities for advertisers to use. What I’m suggesting is that digitally-served (and cloud-accessed) books create an ad medium that can be used just like any other digital ad medium.

Just to be clear, I never suggested (or implied) that advertising would be the dominant model for books, just an important one, especially to authors. Not to Stephen King or to anyone else targeting a million unit bestseller (Seth, this goes to your point). Instead, it’s a way for relatively unknown authors (which are most of them) to get some revenue off their books — any number over zero is incremental revenue for many of them. Their hope, of course, is that using an ad model they’ll attract enough readers (because there’s no barrier to reading the book since readers don’t have to buy a book from an author unknown to them, they can just read as many pages as they want), and that those readers will serve as sufficient evidence to a publisher to pick up the book, publish it, and promote it.

So the real value of an ad model is not in replacing the revenue from selling books, but it’s in freeing up the written word: more authors can publish (most of them will be terrible, I’m not pretending that democratizing media leads only to brilliance, just as political democracy occasionally encourages people of questionable intelligence to run for office), more of their works can be sampled easily — and shared, too, which is something the industry is unwilling to provide under existing business models — and more revenue is preserved.

Seth, yes, I have great proof of how people acquire books. It’s all located in my Forrester report on the eBook market ( For those who can’t access it, I’ll summarize Figure 1, which shows the results of a survey of 2,785 people who read at least one book in the prior 6 months. We surveyed them in May/June of this year. The top two ways they acquired the book(s) they read in that time were: 50% a loan from a friend, 38% from a public library (I know, cool, right?), and thanks to rounding, #3 is also 38%, buying a book from a chain bookstore (e.g. Borders, Barnes & Noble). We separated other types of bookstores, all of which fell bellow chain bookstores. In sharing this datapoint with several of the top 5 publishers over the last six months, they have all acknowledged this fact and one even teased about how they need to sue libraries like the RIAA sued Napster.

Hal, I’m going to raise two points relative to your arguments. The first is agreement: when I taught advertising management at Syracuse in the 90s, I was the rebel in the department because the senior-most faculty member, a legend in the field, had just published a book “proving” that advertising was effective. I questioned all of his data, his methods, and rather annoyingly explained to all my students that he was wrong. But I couldn’t conclusively prove that he was wrong because what was lacking was the measurement necessary to identify the waste. Now that we have more measured media, we are seeing advertising fall because advertisers are more willing and able to cut out the fat. But this is actually progress for advertising, not the end of advertising, as you appear to imply.

This leads to my second point, a disagreement: the media aren’t struggling because of shrinking advertising dollars, they’re struggling because of shrinking direct sales. CDs generate less than half the revenue they did at their peak, magazine and newspaper subscriptions are in freefall across the board, and DVD sales fell last year to produce less than box office receipts generated. Books have yet to fall, but are about to. None of this is because of advertising, it is all because digital media simply don’t cost as much to produce and distribute as analog media. As long as one producer in a market chooses to price content on the basis of digital costs, all producers will have to, putting downward pressure on price and reducing revenues. That’s problem number one with media. The lesson we learn from it is simple: digital experiences generate less revenue than analog ones, even though they are consumed more frequently and at greater scale. I say it this way for a reason: the same is true for advertising. Digital advertising, though consumed more frequently and at greater scale, will generate less revenue than analog advertising. This is a fact of digital life and so when you cite a decrease in ad spending, it doesn’t shock me or undermine my already shaky faith in advertising, it actually fits the model. Plus, it strengthens my recommendation that media providers should diversify their revenues and free their content by including (not depending on) advertising as an additional source of revenue.


And Seth, I love the massage joke, assuming you know it’s a joke because massage is (for now) a completely non-digital experience. There would be no mechanism for automating, delivering, and measuring advertising in that context. Books have long been similarly individual and isolated experiences. But that is changing thanks to Google — that is my point here. Trust me, when massage becomes digital (when they invent Xbox Kinect II which includes force fields that can massage you virtually), I guarantee you there will be advertising involved.


“There are plenty of forms of interaction (like massage, say) that aren’t ad supported, because they’re not mass. And individual books… not mass, I think.”

Seth, that reminds me of a point the writer/illustrator William Rotsler used to make: Censorship is generally directly proportional to the number of people who can observe a given work simultaneously. Thus leaflets, books, small-run newspapers generally fly under the radar, even in relatively restrictive societies.

Two interesting corollaries there:

* It’s amusing to see censorship and advertising behave in parallel circumstances.

* The fragmentation of simultaneous audiences — whether through cable channel choice, time-shifting DVRs, the internet, or what have you — is having interesting effects. (That fragmentation may also be why political ads [and thus campaign costs] keep rising — you just can’t get enough of the electorate all in one place any more, so you have to seek out/buy multiple channels of contact where one or two used to do.)

Seth Godin

Two questions for you:
1. do you have a source for your claim that the main source of books is loans and libraries?
2. have you done the cpm analysis of how much one would need to charge for ads to make it pay for anyone except Stephen King? A book that gets read by a million people is a MAMMOTH bestseller–and with a CPM of $5 (which would be a lot online these days), that’s a grand total of $5,000 before everyone takes their share.

There are plenty of forms of interaction (like massage, say) that aren’t ad supported, because they’re not mass. And individual books… not mass, I think.


“Qualitatively I feel that ad as percentage of GDP should decrease over time, mainly because the measurement of online ads shows that ads are not nearly as effective as was assumed when results couldn’t be measured as well in other mediums.”

Ummm… Isn’t that what I just said? “The measurement of online ads shows that ads are not nearly as effective as was assumed…” Check. Here was my version: “Who more than Google has the data that shows ads don’t lead to sales any greater than the margin of error?”

The increasing accuracy of measurement of ad effectiveness through the use of online media is what’s making the empirical case ads have hardly any effectiveness at all, and are thus a waste of shareholders’ money. This leads to declining ad buys, measured both as a proportion of the economy and in absolute dollars, as ad buyers increasingly realize advertising doesn’t have any rational justification.

There will be some holdouts who continue unsuccessful business practices with the attitude of, “we’ve always done it that way…,” but that won’t last forever. Or even 30 years, at current rates of decline.


@James, agree, the model is inevitable.

Would add this will be a boon to authors. The traditional publishing model doesn’t work well for all authors; this will provide another opportunity in which readers are more likely to sample their content (minimize up front cost to the reader) and that’s a critical first step for an aspiring author.

@Hal: do your numbers include online ads? Qualitatively I feel that ad as percentage of GDP should decrease over time, mainly because the measurement of online ads shows that ads are not nearly as effective as was assumed when results couldn’t be measured as well in other mediums.


” The only reason book advertising has not happened is that the economics of distributing books have required that people pay for them…”

Um… No. You’re showing your age. (Young-ish.) Paperbacks in the 1960’s and 1970’s frequently had a color ad page pasted in. Trouble was, those paperbacks competed in the marketplace against ones that didn’t have ads, and sold fewer copies.

“(A)dvertising is essential to the success of nearly all media—analog and digital.”

Which is why nearly all media — analog and digital — have been having such a hard time lately being successful in the aggregate.

Advertising is dying.

Ad buys peaked as a percentage of GDP in 2002 (at 2.52%). Today they stand at 1.8% — a 28% drop in 8 years. Absolute dollars for ad sales peaked in 2008, and haven’t recovered. If you want to know why newspapers, radio, billboards, commercial broadcast TV, basic cable, and magazines are all having trouble, that’s why.

Yes, Google and others have picked up some of the slack through selling online advertising. Trouble is, it’s not unlike music. Just like digital sales on iTunes haven’t made up for the aggregate loss in the sales of CDs for the record labels, Google’s pickup hasn’t made up for the loss of sales in offline media.

Actually, far from opening the door to an ad-based model when it comes to books, this strikes me as yet another attempt by Google to find something, anything other than search-based ads to make money (ChromeOS and Android, anyone?). They have a limited window of opportunity before their market collapses completely, and they know it. Who more than Google has the data that shows ads don’t lead to sales any greater than the margin of error? And if ads don’t generate sales, how do you justify the money spent on them to shareholders?

If you haven’t read Bob Garfield’s “Chaos Scenario” pieces in Advertising Age, or Eric Clemons’ “Why Advertising Is Failing On The Internet” on TechCrunch, you really should.

But, really… Advising book publishers to pursue ad-based models at this late date is not unlike telling someone in the 1910’s to go into the street cleaning business because horse manure will always fill the streets.

Joel Downs

I fully agree that this is an inevitability. There are already books online in HTML format that have ads around them, but reading a book as a webpage is admittedly clunky. As e-readers and tablets proliferate, publishers will adapt to all available business models for the medium, one of which is certainly ad-supported reading.

That said, some publishers will fight the model, just as some movie and TV show producers don’t want their shows published on the web supported only by ads. Ultimately, the rate of adoption will be a function of the monetization: publishers will run tests with a few titles, and if they can monetize ad-supported e-books as well as they can for-purchase e-books, they’ll move to the ad-supported model. If the tests we’re seeing with TV shows and movies are any indication, the movement will be slow.

James McQ

I’m very impressed, Mr. Joneses, that you are firmly prepared for the future (frankly, most paidcontent readers are), but believe me, I have had this unpleasant conversation many times and expect to have it many more times.


Oh the unintended consequences. Here your reading a book about the famous female Ape researcher and on the top is a flashing Punch the Monkey ad.



Amazing! Ad-supported content! Revolutionary thinking!

Stop patting yourself on the back. Google may do it, but its not the first time people have mentioned or thought about this, and certainly not that revolutionary to create anger.

Comments are closed.