In this week’s talk from Tech Forum 2015, Kobo’s Nathan Maharaj dives into all that data our ereaders have been gathering on our reading habits, and offers some insight into how price points affect reader engagement.
If you’d like to see the presentation slides from his talk (there are charts aplenty!), you can find them here.
You can also find us on iTunes, Pocket Casts, and TuneIn.
(Scroll down for a transcript of the conversation.)
Transcript
Nathan Maharaj: We know that a price-motivated customer isn't necessarily a motivated reader. But that's kind of probably okay because it's more than likely that that's kind of how books have always been.
Zalina Alvi: Welcome to the third episode in our series of Tech Forum talks. I'm Zalina Alvi, the Community Manager here at BookNet Canada. It's no secret that everyone at BookNet loves data. Specifically, though, we really love the Insight data can give us into the habits of book buyers and readers. So we were pretty big fans of this talk from Nathan Maharaj, the director of merchandising at Kobo on the relationship between price points and reading habits. Thanks to the digital publishing revolution, companies like Kobo can now track things like how long you wait to read a book, how often you pick it up, and if you ever actually make it to the end of "The Goldfinch." Of course, all that results in a whole lot of data. But Nathan has done a great job of distilling it into some really interesting trends and case studies. If you'd like to see the charts he references in the talk, you can find his presentation slides at slideshare.net/booknetcanada. Now, here's Nathan.
Nathan: A year ago, when I gave that talk on this new kind of way of looking at pricing, I was kind of done giving the usual talk about pricing, you know, here's a histogram, here's a demand curve, optimize for revenue, try to maintain a state of alertness. You know, publishers were already used to using our portal insight, getting next-day sales data, changing prices dynamically, really engaging with the business of using price to optimize, you know, net revenues. So, you know, that was kind of done, and that was kind of sold. So as an evangelist on that, my work was done. But there's a new kind of conversation to have. It was something we hadn't really talked about. And it was what kind of value are readers getting for the money they spend on reading? Curious question, how do you measure it, right? So, we focused on time. Time became the critical piece. You know, what does it cost to buy and read books was the overarching question. What kind of leisure activity is that? What's your hourly rate, for example? Happens to be that as a digital bookseller, Kobo was really well suited to investigate this. So, here's how it works. You open your device, you load an app, you're reading, you're connected. That device is... You know, for the purpose of creating this magical experience of being synced everywhere, you have to keep sending bookmark data of the reader's here, now the reader's here, the reader has opened the book, the reader has closed the book. So, we gotta fill that and we gotta fill that fast and make sure that your experience is always magical. What that turned into after a couple of years, we realized, wow, that's a lot of data, and it's a lot of timestamps, and it's a lot of stuff that would actually tell you if you looked at it, how are people engaging with books? So, we've done that, and it is a lot of data.
So we wrangled it, we started slicing it in different ways to figure out how people behave in the aggregate. We want to answer some age-old questions, like, you know, as we bump that bookmark along through the book, how long does it take to read a book? You know, how long do people tend to read at a stretch? And how does it all break out by category? It's a whole bunch of questions. And, you know, measuring time is a key way to do it. So, we focused on a small set of metrics, right out of the email marketing and online retailing playbook. So we looked at open rate. Just as you would look at open rate for an email, we looked at open rate for books. Do people actually open the books they buy? More on that in a moment. We looked at completion rate. Do people get to the end of books that they open? And we as readers, we all know that, yeah, mostly, unless it's bad, in which case the book gets fired, you know, or whatever. Like, we all have very deep feelings about finishing books. And it comes down, in some cases, to religious-philosophical conflicts, which we don't wanna get into right now. And we looked at how long they spent with the book. How many sessions did they engage with? And what do those sessions look like in duration? And then we did things like this. So we looked at... You know, let's take a handful of award winners, right? Books people really wanna buy, they're celebrated. These are, you know, the greatest things to read in the land. So here's a cluster of Canadian award winners from last year. And we looked at, you know, what was the open rate on them. And it's a wide swing. And the first thing you notice is the absolute best one is still not 100%. People buy books and they don't open them. This was true in print. It's still true in digital. The notion of acquiring for your future self that has more leisure time than your present self, that hasn't gone away. We still believe our future selves have more time for leisure than we do.
And that's wonderful because dreams are important. But there's this massive, massive swing. Remember, these are award winners. You got one in five people who bought this actually got around to like cracking it open. So they felt well enough disposed towards it to plunk over their money but when it came down to spending their leisure time, they found something else to do. Fascinating, right? Scary. Is anyone scared? Show of hands. Anybody scared? Good. Pain means you're learning. I learned that a long time ago. It was painful. And then we go into completion rate. Okay, so fine. Among those who open the book, what happens? And then we see a big swing here, you know, some people just not getting to the end of stuff. Some people just absolutely trundling along. Still, you don't get 100%. You know, two-thirds is a good performance. And, again, a wide landscape of things, the data as we found it, remember, not as we tried to influence it, just what we found when we measured it for the first time. And then you get average sessions to complete. So you get little stories like, "This one had such a low completion rate because it's really, really long." So, you know, a bunch of readers fell off here. And, you know, these two as similar as they are, people felt really well disposed, they opened them, they completed them at the same rate, but this one was a lot longer. So when it comes down to which book was stickier, which was the more a compelling epic, the prize goes here. You know, no fault... You know, not to say this one should have padded the page count. But you get, like, nuances like that in performance. Fascinating stuff, right? So, we sliced it another way. We took price into consideration and looked at a landscape of readers getting different degrees of value from the money they spent on books. We saw, you know, given the average sell price of top-selling categories, the cost per reading hour, looks like this.
There you go. There's your cost per reading hour. So, given this is your average sell price, literary fiction, cost of premium, biographies, cost of premium, hardcover, "New York Times" type nonfiction, cost of premium. Genre fiction is more of a deal. And as far as the cost per hour that you get out of it, the frugal choice is romance, friends. The budget-conscious reader is reading romance. These spendthrifts over here with their Gladwellian big ideas, as much as they're reading about economics, they're not practicing any of those principles in how they conduct their own finances. We also see a massive fall-off rate. So part of the reason these big ideas folks are paying so much is they're giving up. You know, the suspicion we've had, those books are too damn long, that magazine article got puffed into a book, it's still a magazine article, but there's a bunch of crap hanging onto it. We're fooling nobody. And readers are giving up on this. So, you know, if you've got an at-risk category, something that, you know, the 18-minute TED talk will absolutely assassinate, this is the category. Because the book is only worth an 18-minute TED Talk, in many cases. Okay, scared, more scared, less scared? At this point, just sort of like in vertigo. Great.
So we learned a few things. We learned that less than 100% of readers open the books they buy. We learned it for a fact, not just intuitively. We learned that less than 100% of books that are opened are read to the end. We already knew that, but it's nice to confirm things with data. And we learned that big ideas are really expensive. And they come out to about an evening's entertainment, which is usually what they're for, right? So you can dominate the...you can mansplain your way through a dinner party, I guess is the point of that kind of book.
So, we were left with a few curious points we wanted to explore further. And we wanted to know, you know, how does promotional pricing factor into this? Does that change how people behave with stuff? If you buy the exact same book as the person next to you but you bought it during a sale and they bought it at full price, how does that impact how you deal with that book, how you regard it? You know, I look at all the hardcovers on my shelf, which look quite impressive. But they all came to me through the remainder bin when I was working part-time at Chapters. And I know that. And I have very mixed feelings about them because they look so handsome. And I know I paid 4.99 for many of them.
So anyway, we looked at promotional pricing, dynamic pricing, and promotional activity. So if you were reading a medium article, please return. I'm using the power of vanity to bring you back. There's a picture of you on the screen now. And okay, so new talk is starting now with credit to Kevin Ashton for both the setup for this talk, which I did not pay him for, and the picture. So a quick poll of the room. A show of hands, who thinks the price that a reader pays and their likelihood to open a book are positively correlated? So high price means a high likelihood open. Like the Freud thing, right? Pay for the therapy so you value it. Yeah. Okay. Who thinks it's the inverse? Who thinks you pay a lot and then you're aspirational? You're like, "Oh, 'The Goldfinch.' Absolutely. I'm the type of person who reads 'The Goldfinch.' There's my 20... This is me buying 'The Goldfinch.'" And then you don't get around it, right? So show hands-on case two. I love you guys. You're so cynical. And who right now is having maybe a mild panic attack, willfully abstaining for either philosophical or religious reasons?
Okay, we're gonna have a group ready. Later, we're gonna have a sports circle. It's gonna be a safe space. We're gonna talk this through. It's gonna be fine. Right? It's just data. It can't hurt you. All right. So here is open rate by sell price in $1 increments. All right, Marvel at... Because it's a calming blue. This is important, right? I was thinking about you all. So what we see here is there's this weird thing and promotional activity where absolutely, yeah, if people will get something for a deal, it takes a chunk out of the likelihood that they'll read it. It's not huge. It doesn't, like, completely destroy anything. And by not huge, I mean, it takes them down from the roughly 50% that's normal for everything else. So yeah, it's kind of a coin toss. It's kind of a coin toss. If people buy a book, will they open it? Maybe. It's about as good as we can do. That's the firmest maybe you're gonna hear. And, you know, then it kind of levels out to about 50% through here and maybe drops off slightly here, but not in any way that I think is significant.
So yeah, right? Crazy. Okay, so when people buy cheap stuff, you know, the myth of the hoarder might be true. It might actually be a behaviour that people exhibit. Yeah. And at the same time, though, there's a little bit of a drop-off in the higher price bands. So both groups are kind of right, except for the panic group because we should not panic, it's fine. It's just data. But groups A and B, yeah, the higher price does see a bit of a drop-off for open rate, and the lower price also. So there's kind of a happy thing going on here where people who pay about 10 bucks for a book have the strongest likelihood of opening it. And here, it's weighted by unit sales. So the most intense blue, the most deeply calming blue, is where all the unit sales go. So, what you see here is also the higher price thing is actually fairly insignificant, because there's just not that much action going on there in terms of unit sales. And then there's lots of promo stuff going on here. And here's this mushy middle of, like, mass-market type pricing. And then you get into the, you know, frontlist here with a bit of a stretch up to there.
So, okay, that's open rate, right? So, if people get things for a bargain, they may or may not open it. If people pay a lot for it, they're a little less likely to open it than if they paid a lower, more conventional price. It's not so bad. Here's completion rate. So among those who opened, who gets to the end? Those bargain hunters, if you can get them to open a book, well, this makes sense, right? They don't wanna buy another book because they're cheap. So let's stick with this one. We haven't actually observed that, but that's a question I wanna pursue is, you know, if we cohort the customers who buy down in this range, what do they do while they're reading? Do they, like, tunnel away and we don't see them again until their e-reader runs out of batteries and they have to plug it in, then we get those bookmarks, and we see what they're up to? And so, I don't know, they may disappear entirely. So, yeah, they... So really high completion rate down here. And fairly low down here. Now, that shouldn't mean, oh, crap, nobody finishes books that are $20. There's some nuance here.
Down here is also where you see serialization, right? This is where you see the novellas. This is where you see, you know, the one-hour-long erotic serials, which, for reasons of narrative structure, it's really important to get to the end. If you're gonna start it, you should get to the end. It's the best part. So that happens down here. It would be too obvious to go for a drink of water right now because it would betray a certain level of discomfort, so I won't. I'll put that off till the next slide. And then completion rate kind of levels off and falls down here. And when you get up here, you know, don't fall. These people, they're doing one of two things. They're buying a box set, so, you know, it's the complete "Game of Thrones," and nobody... You know, cast that stone if you're feeling brave for someone not finishing "Game of Thrones," because we all live in a glass house built by George RR Martin there.
And, you know, like Ken Follett books, right? You know, life is only so long. So you got, you know, your Neal Stephensons, your Ken Folletts. Really epic long stuff here, as well as box sets that naturally, you know, people shouldn't get to the end of quickly. You know, you should give someone five years to read the complete "Game of Thrones" box set. That's fair. Stop judging. I can sit on Goodreads for as long as I'm reading it, and that's not a failure. And then you weigh it by units. And you see, in fact, there's not a lot of box sets and expensive stuff being sold. So, you know, the old pricing wisdom is true that stuff tends to sell on this cluster down here, and you've got... And then the cheap stuff is here. So, like, the novella business is healthy. The serialization stuff is healthy. And that gets completed, and that kind of makes sense, right?
So, relieved? I was relieved, because what I rolled up from this is aside from, like, people aren't reading as many...you know, people aren't reading the books that we thought they were reading, they're certainly buying the books. And, you know, you don't get a receipt every time somebody reads the book unless you're in a subscription service, and you can bear that in mind. So, it's kind of a relief to see that people are actually reading books, that the stuff they're tending to buy, where the sales cluster tends to be what they actually open. It tends to be what they finish. So, you know, to build on, I think what something Brian said was, you know, "Good job publishers, looks like you know what you're doing." The stuff that's selling the most is actually getting engaged with. So, you know, engaging content, not a problem. It's being made. This is truly an exercise in optimization. So there was another figure we wanted to look at, though, besides open and completion, because it's a missing time figure. How many days passed before a book gets opened?
So days to open is what we call it. And we look at this as a median. Because, you know, the question is, if a book is gonna get read, how long does it have to hang around on that nightstand before that happens? And that's very, very curious because we often strategically price for the purpose of stimulating a market for something that's gonna come next, you know, knowing only about half the people are gonna read it. And if they get it on sale, maybe a lot less than half. And then completion is another thing. But when does that open rate...you know, when does that figure take place? And here we go. So here's median number of days between purchase and open. Way down here, if you pay a buck for a book, you're probably gonna let it sit for a while. Maybe you feel bad about it, I don't know. That's between you and your God. But then as you get up here, you get a stronger tendency for people who actually read things, you get into the serial stuff, which you're absolutely gonna snap up right away. And then it kind of levels out. And if you're reading the y-axis, yeah, it levels out to about a month. About a month. So like a fine cut of meat. There's a certain hanging in the air, an aging process if you will, to the purchase book. It's not ready to be read right away. That's for the uncultured. That's for the unsophisticated, the people who just can't wait. Real readers know, in the aggregate, you should wait about a month. But then you get down to the higher pricing, this really steep fall off of, you know, if you paid a lot for it, you're gonna get right in there, the stronger tendency to engage with it. Curious, I don't have a lot of hypotheses around that but we do have, again, the unit waiting and, you know, there's not much to hang a hypothesis on here. There's a lot less unit sales. And way more of the action is happening, of course, in that comfort zone we've got right back here.
Okay. So, with the aggregate data, as interesting as is, as fun as it is to kick around, you know, it only takes us so far. What gets really fun is when we drill down to, you know, a title here or a title there and see how it tells a story. Picking up on some of the higher volume stuff that actually saw some activity as a price promo. So, to round this out, we're gonna look at three different books that participated in some form of promotional pricing activity. So, all right. Now, obviously, there's sample bias here. This is not a scientific conference, so I can do whatever I want as long as it's entertaining. That's the only imperative and has to be true, I think. True. Yeah, it has to be true. And we talked about this. No one spoke to me about this.
Okay. So, here's how this book lived in the market. This is book number one, anonymized, I'm not gonna tell you what it was, not even gonna be hints. I'm probably gonna use the wrong pronoun or not to talk about the author. So, it was 12.99. That was the regular digital list price. It got marked down to 2.99 quietly, just happened. This is something that the publisher did for the market, 2.99 is the price. And we didn't do anything special with it. We just left it alone because we had scheduled it to be featured as a daily deal. For one day, we would shine a spotlight on this book alone and drive unit sales for it. Following that day, it carried on at 2.99 before returning to full price. Here's what happened. So that's how long people waited to open it. That's your median days to open. So the bargain hunters, if they were gonna open it, they got to it right away. And right away being in about two weeks, or, you know, a week and a half if they're daily deal people.
But if they bought it full price, it was the normal approximate one month. So, that's interesting but it's not terribly shocking. A huge difference across open rate. So the likelihood that they were even going to open it. So leaving aside, you know, once they opened, it was at this point in the calendar, the likelihood of opening it was far, far higher among people who paid full price. And they hit, you know, that golden 50%. Curiously, though, these people, all three groups, all these guys down here, all paid 2.99, but the daily deal people had an extremely low tendency to open this book, even though if they opened it, they waited about the same amount of time. The tendency to even do so was really, really low. So, again, this hoarding behaviour thing is real, or it seems to be.
And then we look at completion. So, you know, was it a good book? It seems to be more or less flat, right? It's kind of mid-30s, and then kind of 40. So, maybe paying almost 13 bucks drove you a little harder to get to the end. But, you know, we could try this again and see a different result. Yeah, so the quality of the book, once you're in it, you're in it. And it feels like the quality of the experience you're having as a reader kind of takes over. But whether you're gonna get into it seems to have a lot to do, in this case, it did, with what you paid for it.
Now, before you pull all your books from Daily Deal promos, please don't get on the phone with my merchandisers and get on about that. I need to remind you or let you know that less than 25% of the units in this case actually sold at the full price. So, as wonderful as it looks in full-price land, full-price land represents a fraction of the total. So, we're into a game of funnel filling, right? You know, if we know X percentage will go on to read, X percent will actually become in some meaningful way part of a book's, an author's readership, the question is, what do you wanna put in the numerator? Like, how many people are going to load in there?
Sorry, that denominator. It's funnel loading, You know, we're back to classic e-commerce metrics. So, you know, at daily deal, these markdowns is a great way to just get a lot of readers you probably wouldn't have gotten otherwise. Certainly, there is a lower quality of engagement but it's lower quality of engagement versus zero engagement. So, this is kind of the math and practice behind solving the attention problem, using price to do that. Okay. Book two. So this was a true one-day daily deal on a book that was 14.99. It was at the peak of its buzz. Everyone was reading this. You asked them what they were reading, they actually wouldn't tell you what they were reading. They told you they felt they should be reading this. It was marked down to 2.99 for just one day, and featured, of course, as a daily deal. So we drove as much sales as we could for that period. And here's how it shook out.
So, if you bought it at 2.99, you let it sit for a month, you were like, "I shouldn't read that." And you grabbed it for $3 and then you kind of parked it. If you bought it at full price, though, you engaged with it much faster. You chopped that in half because it was the buzzy book. You probably finished what you were reading and then read this next. And that's if you paid full price. Now, what was the likelihood of you even doing that if you were going to open it? At 2.99, far lower. You know, a one in five chance you would actually read it in the first place after waiting that month, but a much higher chance, better than 50% chance that if you bought it full price, you were actually gonna get around to it. Now, completion rate on this one is interesting. This was a long book. Fairly flat, but a little bit higher for the people who paid a little. I don't know why that is. I mean, it might be because, you know, every time they're like, "Am I still reading this?" Do they remind themselves, "I paid $3 for this. This is awesome. I'm reading a book for a month and I paid $3." And they pat themselves in the back.
Whereas the 14.99 people are like, you know, "I think I can be having more fun. I haven't seen my family in a while. I should go outside." And maybe they disengaged because they're questioning how they spent their resources, which were considerable compared to the cheap people. I don't know, hard to explain. I mean, it's roughly flat, but again, the difference is interesting, especially knowing that the book was so long and kicked off a lot of readers. It was a lot of abandonment on this one.
Book three. Starting price 11.99, promoted at 4.99 as a part of a big price promotion of regular priced books marked down. So a second format type repricing. It's weird to say second format in ebooks. It's still an ebook. It's just a different price now. So 9.99, right? That's kind of this paperback pricing. And it was not the first installment in an erotic romance series. So, there's a certain disposition to bring to this data, as we try to understand what the reader is thinking. So, how long did they let it sit on the shelf? If you got a deal on this, you could barely wait a week if you were gonna open this. If you paid full price, you waited considerably longer, but you still did way better than the normal one month, you came in more like three weeks, which is impressive, I think. And open rate, again, we see that spike to full price. You pay full price, you tend to open. However, we should note that the open rate on this is crazy high. It's between 70% and 80%. Could be for a number of things. It's not the first volume in a series, right? So, this is a reader who really knows what they're getting. And if they're gonna buy it, it's because they've got a certain set of expectations that have been built up by earlier books they've read. And completion rate, again, pretty flat, which, you know, as we do more of these, as we dig into a lot of the stuff more at scale, I think we're gonna find that the completion rate... We'll find that that bridge between these case by case things and the big pink graph we looked at before, and that completion rate really is kind of flat because a reader who's reading is just kind of reading, and they're having the experience that they're gonna have, and the notion of what they paid for the book is kind of left behind a little bit.
So, what does all this mean? So here's my take, and you can tell me how you got something else out of it or that I'm wrong. Both actually. When a book is opened, it's often a month after it was paid for. So if it's gonna get opened, it often sits on the shelf a good long time. We know that a price-motivated customer isn't necessarily a motivated reader. But that's kind of probably okay because it's more than likely that's kind of how books have always been. Again, we haven't done anything to adjust these. There's been no, like, Kobo campaign of like, "Hey, read the books you bought." It's really hard to make money on that. So, people buy books and they don't read them. We all do this. It's hard for us to, you know... There's a lot of people in this room who find themselves awash in books that came to them at little or no cost. But, you know, even regular people who pay their hard-earned cash, they may not read what they buy, and they're motivated by price to acquire.
Paying more does seem to correlate with stronger engagement. But, of course, the big caveat on that is... Whoops. The big caveat on that is... Thank you. And we're done. The big caveat on that is you can't make more of these readers. It seems like a really hard thing to persuade someone to become this kind of reader. These people are...they are rare, they are precious. You know, they're endangered species that live in a fragile ecosystem and we should, you know, preserve that to the best we can. But I don't think we can necessarily make more of them. All we can do is, you know, optimize your business around serving them so that they stay happy, they stay reading, and they find video games confusing and frightening.
Zalina: Next week, we've got our very own, Noah Genner giving a lowdown on our latest research on the Canadian book market and Canadian readers. If you want to learn more about what we do, you can find us at booknetcanada.ca. Thanks, Nathan, for speaking in Tech Forum, and to everyone who attended or helped put it together. We gratefully acknowledge the financial support of the Government of Canada through the Canada Book Fund. And, of course, thanks to you for listening.
