Couple holding hands
Building a closer relationship (Image By 日:Muramasa via Wikimedia Commons)

As Kitchen readers know, I have been working on a project on direct-to-consumer (D2C) sales for university presses.  During my conversations with press staff, I was able to put together a list of motivations for why a press would want to sell direct. After all, this is not a self-evident proposition. What’s wrong with the established channels of distribution? What could a university press (most presses are small affairs) add to the distribution capability of Amazon, Barnes & Noble, Baker & Taylor, and Ingram? It’s rather audacious, is it not, to think that someone would come to the Web store of a particular university press to make a purchase when the shopping opportunities on the Web are already so well established; and as anyone who has ever experienced the customer service (and pricing) of Amazon can attest, a university press may find the competition simply overwhelming.

University press people are not stupid and they know that the road to D2C marketing runs uphill; and no one I have spoken to has even suggested that they would like to cut Amazon out of the picture. As for nobody wanting to buy something from an individual publisher when Amazon is just a click away, well, the facts speak for themselves:  some people are already buying books directly from the presses. My estimate is that about 1% of all university press sales are on a D2C basis; that’s 1% of approximately $320 million or $3.2 million. That’s pocket change for Amazon or even the suffering B&N, but for a university press, that’s real money. We don’t know why those people buy directly from a publisher’s Web site, but they do. It does not seem illogical for a press to look at its sales figures and say, “Hmmm. I’ve got 1% from the Web site: How do I get this to 3%? And if I already have 3%, what’s the strategy to get it to 10%?” And that’s the order of magnitude we are talking about here, sales in the single digits, with a low double-digit aspiration on the distant–very distant–horizon.

One easy way for a press to accommodate itself to Amazon and other retailers is to conceive of the site not so much as a retail outlet but as a marketing forum. A user comes to a press’s Web site, looks around a bit, finds something of interest, and then says: This looks like a great book; I think I will buy it at Amazon. This happens all the time, but it’s not clear if a press loses in this situation. The press has already made money selling the book (at a discount) to Amazon, so if the press’s own site functions as no more than a marketing platform, what’s not to like? Indeed, all presses (probably virtually all book publishers) see their own sites as the one place where they can tell the whole story about their books, and if someone prefers to buy the book elsewhere, so be it. In such a situation, the press’s own ecommerce capability is more to provide a user with an immediate option than to build a large sales channel. We can call this “pragmatic D2C,” and encourage all publishers to engage in it.

Some presses report that a major reason for creating D2C capability is the need to diversify sales channels. This is most likely an indirect swipe at Amazon’s market dominance, which is the principal obstacle to retail diversity. It’s easy to imagine a press’s thinking on this point. A sales director is looking at sales figures week after week and sees Amazon’s market share growing. At the same time Amazon begins to demand better and better terms. Who in that situation would not want to look for revenue from a different quarter? Channel diversification is also important if a publisher gets into a real battle with Amazon (this does happen) where Amazon refuses to sell the publisher’s books for a period of time. In such a situation the publisher’s books will always be available in at least one place, the publisher’s own storefront.

Not surprisingly, many presses report that an important reason for them to sell on a D2C basis is that it could improve their margins. I am not entirely persuaded by this argument, but here is how it is supposed to work. A press sells books for the most part on an indirect basis–that is, through sales channels–at a discount off of list price. The discount varies by a number of things, but ranges from a low of 20% to a high of sometimes over 50%. If a press has a monograph with a suggested list price of $40, it may be selling that book for, say, $25 to wholesalers. But on the press’s own Web site the press has the potential of selling the book for full price, that is, $40. That’s quite a big increase in margin, most of which (all but the increased royalty) drops to the bottom line.

The reason I am not entirely persuaded by this is that those presses that report the strongest sales from their Web sites are also the presses that make a point of running special promotions. So a press may run a “spring sale,” offering 20% off the list price for all its books. Now 20% is better than 35% or 50%, of course, but it’s less than full price. On top of this are the hidden costs of handling. A print book has to be picked from the warehouse and mailed to the customer: that is not free, but it’s hard to determine just what that costs. The improved margin for D2C sales gets cut down a bit under close analysis.

The margin argument is an important one because it is one reason that we don’t have an all-university-press online bookstore today.  I have been trying to gather support for such a project for several years, but the presses I have spoken to both want it and don’t want to give anything up to have it. If you had 130 presses (the approximate membership of the AAUP) all participating in an online bookstore, from where would the books be shipped, from a central warehouse or from the warehouses of each individual press?  Some presses would like to ship their own books (under the assumption that their service level would be sufficient to satisfy their customers), but what if a customer wanted to purchase 2 books, each from a different publisher? Then the books couldn’t be shipped together and the customer would incur two shipping costs. Nor does this problem go away with ebooks, as they still have to reside on a server or multiple servers somewhere. To have an effective online bookstore for all presses means giving up margin. I think that’s a small price to pay for an enhanced marketing presence on the Web.

Where most presses agree on D2C sales is that selling direct is a way to establish a relationship with customers and readers, something that is very difficult to do when selling through third-party channels.  The question then becomes what can be made of this relationship?  And on this point there is a wide range of opinion:

  • Collect user data for a marketing database. If I purchase a work of anthropology from the Web site of the University of Colorado Press, the Press may keep my name and then solicit me via email for a forthcoming work of anthropology.
  • Assert Branding. This may sound somewhat abstract to people who are not marketers, but one of the frustrations of being a publishers is that very often customers don’t know the name of the publisher of a particular book. On the other hand, when a customer buys a book directly from the publisher, the customer is made to focus on the publisher’s brand and to associate it with books that are appealing to that individual.
  • Use sales data to improve marketing activity. When a user comes to a Web site, the path of the user through that site can be instructive. Is the site constructed in such a way as to facilitate sales? What kind of marketing messages yield the best results? None of this information is available to a publisher that only sells books through third parties.
  • Use the feedback from D2C marketing to improve the editorial program. A D2C relationship will provide a publisher with insights into the books it publishes. Are these the “right” books? Are these the “right” authors? Are the books too long or too short? Editors work on a combination of cultivated instinct and a network of advisors, but data is data, and D2C marketing generates data that few publishers ever see if they don’t attempt to sell books directly from their Web sites.

While I was researching this project I interviewed a senior executive in the university press world who made an insightful remark. As for all this information people are trying to gather from D2C activity:  “What are they going to do with it?” This is not an idle question. It is one thing to collect data, quite another to know how to analyze it and build and implement programs based on it. I cannot shake the feeling that presses have only a sketchy notion of what to do with the data they collect. The direct relationships they seek to forge with their direct customers may be somewhat empty. I am reminded of a story I heard recently about two people who met through online dating. After exchanging messages online they agreed to meet for lunch.  And then they discovered they had absolutely nothing to say to each other.

If presses are going to have something to say to their customers, they are going to have to make a concerted attempt to strengthen their D2C activity. They will have to rethink their Web sites (traffic is the currency of the World Wide Web) and they will have to learn what information about users they should collect and how to use it.  It’s not enough to hope for improved D2C sales; these sales have to be fought for.


Joseph Esposito

Joseph Esposito

Joe Esposito is a management consultant for the publishing and digital services industries. Joe focuses on organizational strategy and new business development. He is active in both the for-profit and not-for-profit areas.


31 Thoughts on "Why Do Publishers Want to Sell Direct?"

What if a customer wanted to purchase 2 books, each from a different publisher? Then the books couldn’t be shipped together and the customer would incur two shipping costs. Nor does this problem go away with ebooks, as they still have to reside on a server or multiple servers somewhere.

I see the problem when shipping physical books. But isn’t the shipping cost of an e-book zero? In which case why is it a problem if the buyer has to pay it twice?

Mr. Taylor:

There is no free lunch even when shipping e books.

There are a number of steps in e publishing the first being making it available in shipable e form.

I could go on but to assume that e books are free when it comes to distribution is beyond the pale.

Nobody disputes the startup costs involved in any form of commerce, whether physical or electronic goods are being sold. What I’m asking about here is marginal cost. If it’s not the case that the marginal cost of creating a delivering a copy of an ebook is zero, perhaps you would explain why not, rather than just saying that the idea is beyond the pale?

(By the way, it’s Doctor Taylor; but just plain Mike is fine.)

Dr. Taylor, (I can never bring myself to call you Mike again), the cost is in the management of the metadata and links to the various distributors. I don’t know if anyone has solved this efficiently yet when the files are hosted on multiple sites, but most online stores use a single site for hosting. It adds to the complexity of the operation to have multiple sites, which in turn means higher costs for IT support. One solution would be to simply put all the files in one place (e.g., Ingram’s CoreSource), but then you have to give up margin.

Please send my regards to Professor Harnad.

Thanks, that makes sense. (Why wouldn’t you call me Mike, though?)

Dr. Taylor:

I urge you to take a course or read a book on the finances of publishing and many of your comments would be answered.

I’m not with you on this one, Harvey. The blog’s comments can be a good place to try on ideas, learn things from people you wouldn’t otherwise get to talk with, etc. Dr. Taylor’s (aka “Mike”) questions are offered in good faith and his level of engagement is inspiring. I would add that most books about publishing make for dull reading. I would much rather read a book for the layman by Dr. Taylor on paleontology.

Joe and Dr. Taylor, I apologize!

I overeacted. I hear so much about with e delivery everything is free that I become frustrated.

But i would recommend Books in the Digital Age (Polity, 2005) by British sociologist (and co-owner of Polity Press) John Thompson, which is anything but dull.

The Thompson book is very good, but it is mistitled. It should be called “Books Right up to the Point before the Digital Age.” Nothing digital is seriously discussed in it.

Thanks, Joe, much appreciated! I’m afraid I’ve never written a book on palaeo; but I can offer One on Doctor Who if that’s any use!

Meanwhile, if anyone can recommend a good book that will tell me all I need to know on the finances of publishing, I’d be happy to hear about it.

Not sure if such a book exists. The quickest way I’ve found is to spend 5-10 years in the industry. I have to say that it took me about the same amount of time to feel comfortable in the molecular biology laboratory, to really feel like I had a clue about how things work, as it did in scholarly publishing as an editor. Sometimes there are no shortcuts.

The Art and Science of Book Publishing written by former Princeton University Press director Herbert S. Bailey, Jr., once used to be that book, but it was written well before the digital age, so would not have anything to say about finances in relation to e-publishing.

I am not familiar with any one book that does the job. There is more about book publishing than journal publishing (the book biz is 15 times larger than journals). Most people learn about journals on the job. The big guys hold their information closely, so it’s hard to learn much without serving an apprenticeship. There are workshops at industry conferences that address some of these things (e.g., SSP in Boston next month).

I have been really surprised by the small share of university press direct to consumer sales that you estimate. I would have thought that a very large fraction of sales were through this channel. I usually buy academic books direct from the publisher and always have (not that I buy many books). They have always put a big effort into sending catalogs out to academics.

Joe, very good post on an important topic. I have to disagree with your comment about D2C not improving margin. True, margins may be the same for the first sale. But you do not attribute any value to the name and address acquired from the purchaser. As you go on the mention further down, there is a great deal of value in that asset. Specialty consumers blisters have been monetizing their consumer databases very successfully for years. When you add in the future value of the names gathered, margins are definitely higher.

The costs of online marketing have plummeted over the past several years. Online marketing is easily within reach for even the smallest publishing house and probably makes more sense than through a centralized store that has the issues you outline so well.

Let us not forget that collecting customer data (which, indeed, is valuable), analyzing it, and acting on it is time consuming and, just like other press activities, requires resources.

Many thanks for your periodic posts on this topic, Joe. This one has prompted a couple of comments for me; apologies if you’ve addressed these very points in previous posts or plan to in future ones.

First of all, one extremely important thing that distinguishes university presses is that in a very real way they often “own” a discipline (or a small number of disciplines). This means that unlike the typical publisher working through channels, they _already know_ who the customers are (okay, not all of them, but a meaningful number, and often thought leaders) for books in the discipline(s) they focus on. And those customers know them, too: their brand is _already_ important; in fact, those customers are also their authors in many cases. So it is not just a matter of cultivating a brand, and of developing relationships; for many university presses it is about _exploiting_ a brand and the relationships that they already enjoy. Many other types of publishers would give a lot for that.

My second comment is just to point out that there is another alternative to either relying on retail channels like Amazon and B&N or taking on the burden of ecommerce and fulfillment themselves (which many university presses are just, frankly, too small to do well). University presses have another option that doesn’t exist for many other publishers: the services that Chicago, Toronto, et al. provide.

By focusing on the disciplines they are known for and by leveraging the resources that are available to them, I think that goal of 10% D2C is much more attainable for many university presses than it may appear to be.

I wonder if the best opportunities for D2C sales may come from regional publishing? Most, if not all, of the presses associated with state universities do regional publishing, and for some it can amount to as much as 60% of revenue (I heard that figure once cited for the University of New Mexico Press). Does Amazon, or B&N stores in the region, stock all of these regional titles? Marketing regional titles also differs significantly from the marketing of scholarly books; the audiences are quite different, with the scholarly market readily identifiable and reachable while the regional market is more amorphous and requires different strategies to reach.

Hi Joe,

Interesting post. I used to run a small publishing company so I might be able to shed some light on this mystery. My experience may not fit the norm, but there are far more considerations involved in the motivation for direct sales than just margin. Here’s the gist of a distributor contract that most small presses are offered from distributors: (1) 55% discount; (2) publisher pays for shipping both ways (to distributor and back from distributor in case of returns); (3) no minimum order limit (so distributor can order one copy); (4) unlimited duration on returns, subject to full refund (even if the returned book is no longer sellable).

Let’s look at just the first condition—the 55% discount. On a short run of books, after factoring in printing, editing, design, shipping, and marketing costs, the unit cost of a book may be $10 and you’ve marked it for retail at $20. So, by fulfilling an order from Amazon for one book, where the publisher needs to pay outbound shipping on top of a 55% discount (and not even including order processing time and labor), the publisher is taking a loss.

Given this and all the other conditions (imagine the accounting chaos that ensues when you get a return 180 days later), there’s a clear motive to sell direct—not only to avoid the distributor discount, but so the publisher can charge for shipping and handling, stipulate that sales are final, etc. Offering a 20% discount off retail doesn’t affect this calculus.

This whole scenario assumes that the publisher’s book (or books) are not big sellers and that most sales are limited and focused—not best-selling paperbacks, but (for instance) business, trade, or technical books with focused audiences who don’t necessarily need to be reached through bookstores.

Hope this helps.


Glenn Hampson

I often think that once things settle down and we think about how to sell books going forward, we will end up pretty close to where we are because it works. I do agree about the marketing aspect. I think the visibility offered by some low margin, even or mild losses should be planned as part of the marketing strategy/expense that yield sales and relationship in the future and in some cases longterm. Single digits are still part of the plus column and if a press can maintain or grow a bit-it is all good. Publishing is a complex business that folks think can be easily reconfigured. Even w all of Amazon’s success, they have low margins on new books, make money on other services, didn’t understand distribution until they were already in business and took several years to turn a profit.

There are two primary reasons why a publisher would want to sell direct. As you suggest, the consumer data collected (not just the email address but the consumer behavior) has value. This value is what eCommerce folks call “customer life-time value.” For a sale that occurs off-site the value is the revenue from the sale; when it occurs onsite it’s the value of every transaction that consumer makes on the site over x-period of time. The second reason to sell direct, which you also hint at, is that the marketing efforts to drive people to the website will yield sales both on the site and off. But more important in a way is that by selling direct you then know what marketing efforts produce the most fruit on site, which is the best metric you can have for what’s likely to generate sales off site as well.

This post and the thinking behind it is exactly right, Joe.

U presses should build a unified web presence; the breadth, depth, and inter-connectivity of the information provided would deliver unique value to visitors (note that I did not say “books” in that clause), and the data would have salutary effects on communication between scholars and all participants in their ongoing work; i.e., improve understanding and encourage new insights.

By now we all should have done some work in database marketing, logit analysis, and/or at least regression analysis to project trends, or at least know of these disciplinary applications enough to understand the complex benefits you are imagining for u press. I can say with confidence that there is no doubt that presses, currently, are sitting atop a wealth of valuable actionable information; were these streams but to simply converge, the benefits to the presses, institutions, and more importantly the “customers” would be sincerely favorable.

But after that site is up and running, have folks thought about what comes next, the new possibilities? What folks need to do to follow you, to see things differently, is “let go of the side of the pool.” Until we do let go of the side of the pool (and accept the author’s donnee), we’re not really swimming; we’re just all wet.

I’ve cheered your work in this area for years; so, you know you have my support. The only thing that I can add, is that that nebulous thing that we sometimes call ‘the future,’ is on your side here as well. Nil desperandum. Keep the posts coming! Thanks again for all your work on this!

Vale, Magister!

For what it’s worth, while I agree with you Peter, I think university presses should avoid at all costs the “big box” approach to direct marketing and eCommerce. In order to succeed in the online book retail environment, one has to deliver value to consumers other than lowest price and ease of ordering. One key way to do that is to create user-facing marketing/eCommerce platforms that are oriented around consumer interests with a backend integrated with publishers and distributors who can provide the necessary title metadata and a fulfillment partner. I’m actually working on just such a platform and backend.called In Books & Media, which is seeking investor backing.

Peter, I’m unclear on what that data might reveal that’s actionable. Would we change our editorial programs as a result, and how might we do that without compromising our mission? Might the end result be an environment where we publish the popular rather than scholarly excellence? And what does that do to the less popular disciplines like romance languages or literary criticism, and the scholars in that field who depend on us?

I use statistics to analyze our marketing efforts, but I’m uncomfortable making the leap in suggesting that we can collect useful information from our own sites to appropriately inform product development—maybe print runs, but not editorial programs. User time spent on a university press site isn’t quite like the analytics one might collect at, and probably shouldn’t be used the same way.

Before anyone lets go of the side of the pool, they need to be sure they have enough information on how to swim, rather than a lot of data on the temperature of the water, and the number, age, income level, and gender of the other people in the pool. It might just make more sense to simply get out of the pool and walk around to the other side.

I love your image of getting out of the pool and walking round it, Tony, but what a decade in the supply chain taught me is that data is always a great starting point for making decisions. The art is in designing the data capture properly and knowing what questions to ask of the data (these two are not un-related). I agree with Joe that D2C web activity seen in the context of the overall marketing activity of a press is the way to go on this, particularly in a world where we accept principles of content marketing, and also that online marketing is a dialogue not a broadcast.
As to the notion of a transition from scholarly excellence to popular, I’m currently doing some work with a distinctly non-scholarly press that does over 50% of total sales d2c via their web site. Lots of what they do would not transfer naturally to scholarly publishing, but what I have seen is that there are tools out there that give fantastic user metrics, and SEO techniques that really work. I think that to get to grips with this, university presses need to face the prospect of employing people from different backgrounds, with different skills who may not take loyal commitment to the scholarly endeavour as read in the same way as most university press employees do. Bringing in outside expertise is a difficult thing for many university presses to do (aside from anything else, pay scales can really work against presses in this regard). However finding a way to make it possible would pay dividends.

I have run my own ebook website (not a University Press) for several years now and I can verify that the four benefits you list – branding, collecting data, using data for marketing and getting feedback – are all extremity important and actionable.

You don’t need to be an expert to utilize this data; just a quick peek at Google Analytics will show you where most of your buyers are coming from, and what ads and articles are bringing them to your site. Focus on those and your sales will grow. This is really important info that a seller on Amazon will never see.

As the author notes, you can still sell through Amazon or other outlets in addition to marketing yourselves. And universities should be an easy place to find smart students who want an ecommerce internship on their resume. Why would publishers not want to do this?

Reblogged this on Exploring Ideas: Shaping Business and commented:
Over at The Scholarly Kitchen, Joe Esposito has been writing up more of his research regarding d2c sales by university presses. Both the post and the comments thread are fascinating and I recommend them both. As I posted here recently, I don’t think that 10% of sales direct by university presses is an unachievable target. I’ve worked with independent publishers who achieve significantly more, and I’ll be exploring the questions Esposito is looking at more in future posts

Comments are closed.