Does library ebook lending hurt book sales? Tor Books experiment reveals answers, may lead to new ebook lending terms

Note: This exclusive report is available free to the public. If you like my reporting on genre issues, including my weekly Genre Grapevine columns, consider backing my Patreon.

I love libraries. I have access to a world-class library system and wouldn’t give that up for anything. So when I read a recent Good e-Reader article with the alarmist headline “Major publishers are sabotaging the public library,” I was outraged. Especially when that article connected with what I’d read in other media outlets about ongoing tensions between publishers and libraries over ebook lending.

Except maybe we’re not getting the whole story from those articles.

Yes, we’re witnessing a sea-change in how publishers and libraries operate in an increasingly digital world. A large share of the readership and revenue for publishers now comes from ebooks, while libraries are evolving into something akin to “spacious multipurpose community centers with increasing demand for computers, digital books, movies and music.” My local library system, for example, is approaching their 18th million digital ebook loan.

On July 1, Hachette Book Group became the last of the big five publishers to end perpetual access licenses for library ebook purchases. Instead, large publishers now require libraries to purchase ebooks and digital audio books under two-year metered terms. This is increasing the costs for many libraries and causing push-back from larger library systems.

However, one aspect of the Good e-Reader report that really caught my eye was Macmillan and their Tor imprint testing an embargo system on library ebook purchases. Launched last year, this embargo test was an attempt to determine whether library ebook lending had a negative impact on retail ebook sales.

To learn more about this embargo test and its findings, and to see if the adverse media coverage I’d read was correct, I interviewed Fritz Foy, president and publisher of Tom Doherty Associates, the unit of Macmillan that includes Tor. Foy provided an unprecedented look at their embargo test and what it discovered.

“The main thing we’re dealing with here is when libraries started ebook lending, we walked into uncharted territory," Foy said. "There are no good guys or bad guys in this story. Instead, we need to better understand how publishers interact with libraries and what is the expectation of consumers with library ebook lending.”

What Tor’s embargo test discovered

Foy said they initiated their test with the understanding that library sales are a major source of growth for the publisher. Macmillan’s library ebook revenue increased 800% over the last five years, with libraries now making up 45% of the publisher’s total digital reads.

However the problem, according to Foy, is that while libraries make up 45% of Macmillan’s ebook reads, those digital-lending reads account for only 15% of their revenue. That readership/revenue gap was starting to have a significant negative impact on Tor’s authors and the company.

“What we were seeing was really reaching a tipping point,” Foy said, “where we’d have to explain to our authors that while your readership is growing, your royalty statement will be getting smaller and smaller.”

To discover if library ebook lending was indeed hurting sales, Macmillan used their Minotaur imprint as a control group and Tor Books as an experimental group. The two groups have books which sold in similar patterns along with authors and book series which drove steady sales from year to year.

For the experiment Tor prohibited ebook sales to libraries until four months after a book’s release. After that date libraries could purchase the Tor ebooks. The control group Minotaur instituted no such restriction. (As a side-note, Foy said the there was never a plan to do a six-month embargo on ebook sales to libraries, as reported in that Good e-Reader article.)

Foy was surprised by the experiment’s stark results.

“All but one title we compared (in the Tor experiment group) had higher sales after the four month embargo on ebook sales to libraries,” he said. “And the only title where we didn’t see this happen had bad reviews. And when you looked at the control group, sales remained the same.”

Update: To prevent any misunderstanding, let me state that Foy specifically said the embargo resulted in higher total sales for Tor for almost every title they tested. The Minotaur imprint control group saw sales remain unchanged.

While Foy did not reveal how much of an increase in sales resulted from the embargo, he described it as significant.

Foy did admit the four-month embargo resulted in lower ebook sales to library systems, which isn’t surprising since most library ebook lending happens in the months immediately following a book’s release. But Foy said this drop in sales was more than offset by the gain in à la carte sales. The embargo also impacted physical book sales, with Tor seeing sales growth across all book formats.

Ongoing discussions with libraries

However, Macmillan didn’t merely run the embargo experiment over the last year. The publisher has also been discussing the larger issue of ebook sales with both the American Library Association and individual library systems.

“We’ve learned a lot from these conversations with ALA and library systems both big and small,” Foy said.

For example, he discovered that the current ebook pricing had a different impact on libraries depending on the size of the system. Most ebooks are currently sold to libraries on a two-year contract allowing that ebook to be lent either 26 or 52 separate times. If a book is lent for a standard two week period of time, 26 separate lends will often cover a year while 52 lends will cover the entire two-year contact.

However, not every library has the same number of lends per ebook. Books on the bestseller lists account for the majority of any library system’s ebook lends, and big library systems have more lends than small systems since they have more patrons.

While these facts may sound self-evident, the big/small system dichotomy can result in different-sized library systems paying different prices for ebooks. If a big system purchases an ebook with 52 lends and lends it out all 52 times, their price may be only $1 or so per lend.

However, a small library system with relatively few patrons may not be able to lend out an ebook for all of the purchased lends. That makes smaller libraries pay exponentially more per lend, up to $8 or more. 

“We sort of backed into terms that were disincentive for smaller libraries carrying our books,” Foy said.

New terms expected by end of summer

Now that Macmillan and Tor are satisfied with the findings of their experiment, they are working on the details of how to proceed from here. While Foy wouldn’t reveal their final plans with regards to library ebook sales, he said they aim to make an announcement before the end of summer.

“I can’t tell you what our solution is because we haven’t enacted it yet, but we’re getting close,” Foy said. “We presented it to three groups of libraries, about thirty all told, and we’ve gotten some good reviews and some mixed reviews but so far no really bad responses. We’re trying to come up with a solution that works with everyone long-term.”

Foy added that authors should also do better under the new plan when it is announced. “Our terms will be seen as good news by a lot of people,” he said.

But Foy emphasized that no matter what the terms may end up being, it’s also important that the relationship between publishers and libraries remain healthy.

“We’re in a difficult place with our authors, so we needed to take a step back and have a lot of conversations,” Foy said. “But we also listened to what (libraries) said and tried to come up with a semi-happy medium.”

By becoming a patron, you'll instantly unlock access to 125 exclusive posts
By becoming a patron, you'll instantly unlock access to 125 exclusive posts