S1 Ep23: The Minimum Viability Threshold (Why Some Books Don’t Sell)

…with Dr. Kerry Spencer Pray

In 2008, Dr. Kerry Spencer Pray watched her book die on submission after the financial crash halted publishing acquisitions. She then decided to do a PhD, focusing her research on what makes books sell.
 
After studying hundreds of factors, Dr. Kerry Pray zeroed in on a handful that seemed correlated with book sales in some way. All of them were marketing related, and often linked to a book’s advance (which is itself often representative of the support a novel gets). Let’s dig in and have a look!

Show Notes

• Dr. Pray talks to us about the data she collected, the methods she used, and the research she produced in pursuit of this project, focusing primarily on the YA market
• End result: she developed an algorithm that could predict the success or failure of a given novel, with about 75% accuracy
• Her other big find: the minimum marketing viability threshold (ie, the point at which a book has a shot at selling well, which is defined in this case as selling within expectations relative to its advance)
• Books can be assigned a marketing viability score, and books which fall below that threshold have zero chance of success (that she has studied so far)
• Most well-marketed books still don’t succeed, but they have a significantly better chance of doing so, and when they do succeed, they sell extremely well
• We discuss various factors and angles of this data / research
• Dr. Pray lays out the eight factors correlated with book sale success
• We discuss outliers, or apparent outliers, and the “myth” of self made success that accompanies very good-selling books
• Sunyi’s theory on why high concept books might thrive better in this kind of environment, and also her metaphor for books as rocket launches (the concept of trying to get into orbit)
• Scott’s speculation on whether publishers run these algorithms themselves, and why or why not they might do so
• The surprising importance of Twitter, and the irrelevance of star ratings
• General chatter, and where you can find Dr. Pray to get in touch

Links

Dr. Kerry Spencer Pray’s university page

YA Book Sales & Book Marketing slides (please do check these out! They’re very good as a starting point for her research, are easy to read, and are a great accompaniment to our episode this week.

Dr. Kerry Spencer Pray’s Twitter (@Swilua)

Transcripts (by Sunyi Dean)

[00:00:01.690] – Sunyi
Hi, I’m Sunyi Dean.

[00:00:03.690] – Scott
And I’m Scott Drakeford.

[00:00:05.830] – Sunyi
And this is the publishing Rodeo podcast. In 2022, we both launched debut novels in the same genre with the same publisher in the same year. But despite having very similar starts, our books and subsequently each of our careers went in very different directions.

[00:00:21.890] – Scott
That pattern repeats itself throughout the industry over and over. Why do some books succeed while others seem to be dead on arrival?

[00:00:30.570] – Sunyi
In this podcast, we aim to answer those questions and many more, along with how to build and maintain an author career.

[00:00:38.410] – Scott
Everyone signing a contract deserves to know what they’re really signing up for. In an industry that loves its secrets, we’ll be sharing real details from real people. We’ll cover the gamut of life as a big five published author, from agents to publishing contracts, finances, and more. Man so I’ve looked through it before, and I’m looking through it again. Your PowerPoint on your whole study, and God, it’s so good.

[00:01:12.870] – Kerry
I’m glad you liked it.

[00:01:16.870] – Scott
Have you walk everybody through it? I want you to just walk everybody through it. But they won’t have the visual. I mean, Sunyi, you can put the link in the show notes.

[00:01:27.620] – Kerry
We could link it if you wanted.

[00:01:29.320] – Scott
Yeah, that’d be a good idea. This is going to be a show and tell episode today.

[00:01:35.020] – Sunyi
I can link it on the Instagram and all the social feeds.

[00:02:25.710] – Scott
Yes. So let’s back up, noting that, god, I have to look around my camera to see my goddamn, I am flustered today. But, yeah, let’s do an intro first. Tell us just who you are and why you ended up doing this and all that. Yeah. Sony. Go ahead. It’s all you, right?

[00:02:45.270] – Sunyi
So today on the publishing radio podcast, we have with us Dr. Kerry Spencer Pray, who I first encountered as Mrs. Frizzle. I’m guessing that’s a reference to Mrs. Frizzle’s magic school bus on Twitter. And she did a brilliant thread about how marketing and sales are linked, where books basically can’t perform sales wise. Outside of their marketing bracket. And I thought that was really fascinating because at the time I was still always hunting for information about how Trad Pub worked and how my own industry worked. And I bookmarked that thread. And then earlier this year when we started our podcast, I circled back round to it and sent it to Scott and I was like, we should see if this person would like to come on and chat to us because I think there’s so much good information there. And he was blown away by the slides and the research he’s done. So if you’d like to introduce yourself, that would be absolutely brilliant. We are very thrilled to have you on.

[00:03:38.150] – Kerry
Well, hello. I’m glad to be here. My name is Dr. Kerry Spencer Pray. I’m an assistant professor of writing at Stevenson University, which is in Maryland in the United States. The thing we’re talking about today is mostly my dissertation research, which the impetus for it goes all the way back to 2008. So in 2008, I had my first sort of fiction book that was on the market. My agent was shopping it around and we were starting to get some interest in it. There was a couple of publishers. One of them started talking to us. We’re talking about acquisition talks and starting to do a little of the negotiations. And then I don’t know if you guys remember 2008, but there was like this massive crash in the entire economy. And there was this day in publishing where in a single day they lost like two years’ worth of revenue or something in returns alone. And what happened? They shut all of New York shut down. It shut down fast and hard and all new acquisitions were cancelled. All new contracts were suspended. They had like everything just like all the doors closed all at once.

[00:04:51.840] – Kerry
And so that book was dead. And my agent was like, well, we’ve already shopped this manuscript. We don’t know when New York is going to be reopening, so you just need to write another book. And I was at the time, I was like, I don’t want to write another book that is hard. Book reading is hard and it takes time and this is depressing. And so I was like, instead I’m going to get a PhD. So I did that instead. I went to the University of Wales in Bangor in the UK. I did that because I already had a master’s degree and I was interested in a creative writing program, but they didn’t really have PhDs and I didn’t want another master’s degree. I thought it would be I was already teaching part time at the at the university that I was was at near those I was living near in Utah. And so I started this PH. D. Program and I decided because I was fresh off this deeply humiliating isn’t the right word. It was sad, it was heartbreaking to have a book and work so hard on it and get an agent and be about to sell it and then have the entire bottom just taken out.

[00:06:05.190] – Kerry
And it was back sort of in the days when Twilight was super popular and my book was different than Twilight, but not like I was Mormon back then. That’s a whole other story. And Stephanie also was Mormon and was in the same circles I was. And I just didn’t understand why was this book doing amazing and why was my book dead? I didn’t understand. And I have sort of a data brain. Like, I studied engineering as an undergraduate and then I got a master’s degree in Early Brit Lit, but I had done a lot of language studies and software things for that. I worked for Oxford’s Canterbury Tales Project, where we digitized and transcribed all of the existing manuscripts of The Canterbury Tales. And then we used Clolidistic analysis, which is from evolutionary biology, to start to look for connections. I had to do a dissertation for my PhD. And so I decided I was going to decide to study what makes books sell or not sell. And it was sort of like everyone was really shocked by this question. It was like this forbidden question why are you what sells or not sells? Like, obviously a book.

[00:07:18.720] – Kerry
If it’s good, it will sell. Or obviously we don’t ask that question. That’s like a debasement of art. And I was like I was writing young adult fiction primarily at the time. And I’m like, listen, young adult fiction is literally defined by the market. It’s not a real genre, it’s a market category. And so you have to understand the market if you’re going to write Ya. And so they decided to let me do it. They were sort of amused by the little chaos gremlin, apparently, that I appeared to them at the time wanting to know why books sell or not. And so I met with the statistician and we came up with a plan and I started a sample. I collected a sample initially of about 200 books, and we set about to figure out why they sell or don’t sell. And I was hoping to find something that would be in my control, like, what could I do that would make my book better able to sell? That seemed like it was something I was interested in. And the sad sort of answer at the end of that was there was literally nothing like, I tested hundreds of book characteristics and there were only like a few that had any connection to sales.

[00:08:32.520] – Kerry
Actually, I think it might be like eight. I have the actual numbers here. Yes, there were eight book characteristics that were correlated with sales, and every single one of them was a marketing factor. There was literally nothing else besides marketing and did a bunch of other sort of tests to try to see if that was really true or not. And then the statistician helped me figure out how do we factor out marketing so that we can see, is there something that is making them sell or not sell after you factor it out? The answer to that was not really. There was one characteristic that we can talk a little bit more that also makes me laugh a little bit, because once you factor out marketing, there’s one that it’s not quite statistically significant, but it’s almost there. And so there’s like because humans are little chaos monkeys, you sometimes have to look at those almost as if they might tell you something. But anyway, the only one that was not a marketing factor that almost had a connection to sales was a quote called, like, the book made me feel Yearning. And it was like a true or false question on our long questionnaire that made me laugh really hard because I write Ya and so it’s all about the yearning.

[00:09:46.400] – Kerry
And that made me laugh. But it was mostly just very depressing because it was like, oh, it’s just marketing. And not only that, and so I continued studied it for a lot of years, and I’ve gotten grants over the years, and I’ve had research assistants over the years and consulted with different publishers over the years because they don’t always understand why things are selling or not selling. But yeah, I didn’t write again for a long time. The next book that I wrote, I wrote a book about a plague, and I wrote it in 2019. And my agent decided to go on the market with it in June of 2020. And nobody really wanted to read a book about a plague in June of 2020. So my publishing of fiction continues to be a very depressing subject, but I’ve published some nonfiction in the meantime.

[00:10:33.290] – Sunyi
I will say when I started out looking at publishing, because my deal size was good, and the first thing I started looking at is like, right, what are all the things that can go wrong? Because I have a doom brain that focuses on all the things that can go wrong. And basically my conclusion after looking at lots of things and talking lots of people, is the book probably won’t go wrong as long as they spend enough money on it. It might underperform for what they want, but it’ll still sell as long as someone is giving it a good shove. I think one of the things that I was interested in when I was reading the thread is you’re talking about how books perform in brackets relative to their marketing.

[00:11:07.690] – Kerry
That was partly just like I think part of that was the way that we had to get the data. So I mentioned before we started recording that I tried to buy data from Nielsen. So Nielsen Book Scan records a ton of data about book sales, about point of sales for all kinds of titles all over. And I had a grant and $5,000 of it was allowed to be able to be spent on data of this grant. And so I tried to buy data from Nielsen, and they found out first they’re like, oh, you’re an author, sure, that’s fine. And then they found out that I was a professor. And they’re like, no sorry, we won’t give you this data. And I was like, Why? And they’re like, you might study it. I was like, exactly, I would like to study it. And they’re like, no, you need to have an industry interest. I’m like, I’m literally a writer. I do. And they were like, no sorry. So they wouldn’t give it to me. And that was frustrating. And so we used a bunch of sort of workarounds. And I actually found, like most publishing people were very cagey and didn’t want to talk to me about this.

[00:12:11.930] – Kerry
I had a couple of editor friends who all like, on conditions of anonymity, they would never tell me their names, and they would never let me speak about them on the records. But what they did do is they would say, okay, here’s some assumptions that you can say or true or false. And so they said if a book has a bigger deal size, like if we spend more on the advance, then it’s going to sell better than if it doesn’t. And that was one of the things they said. And also, like a series. Sometimes what they can do is if they do a three book series, they can take the entire marketing budget for three books and put it in the first book, which makes the first book launch harder, and then the rest of the book sort of borrow off of that marketing so it has more marketing that way. And so they told me a couple of things like this that were not specific and not by house, and not about anything like in there, that they felt like they could tell me. They still even didn’t want me to ever use their names in public or even the places they worked.

[00:13:09.500] – Kerry
At least one of them has left publishing at the time, so I probably could reveal the publisher she worked at, but I won’t after that is I used, based on some of that little information they were willing to give me, I went to Publishers Marketplace, and they have deal classifications of books. And so they have the nice, the very nice, the good, and each one corresponds to a certain number of dollars in the advance. And what that let me do is sort of create five different categories for the advance size. And so I think that’s kind of how it ended up being in brackets that way, like it was marketing score based. And so working with the statistician, what I did is we created an algorithm that first what we did is we measured each one of these things in correlation to sales. And to get sales numbers. That was a little tricky at the time. What we ended up doing was using multiple different online metrics. I say we, I mean, I did this, I did have people that helped me online that read a lot of books. I read about 200 of the books in the sample.

[00:14:13.420] – Kerry
The sample size I have right now is about 500 books big. It’s a demonstrably random sample. And so we took every single Ya deal report that was in publishers marketplace that specified the deal size, or that specified that it was an auction or preempt, or any of the things the editors had said might lead to a higher marketing budget. I did something called a Pearson’s correlation. So a Pearson coefficient, what it does is it’s a way to measure the extent that one set of variables will increase or decrease as another set of variables increases or decreases. So if marketing goes up, do sales go up? So those are like two different sets of numbers. So once you have two sets of numbers, you can see if they’re correlated. And there’s some limitations of a Pearson correlation. Like it can only spot a correlation, it can’t spot causation, so you can sometimes guess. So one of the ones that was a little tricky was there is a correlation, a strong correlation, between the number of Amazon ratings that a book has and how well it sells. And is this a causative thing? And so I’ve seen self pub presses will say, try to get as many people as possible to review it because that will increase sales.

[00:15:34.540] – Kerry
I’m not actually sure that that’s true. I think that books that are selling better tend to have more reviews. So that’s an example of one of the limitations of Pearson relationship coefficient that you calculate. So basically it’s this number between go ahead.

[00:15:48.030] – Scott
Yeah, no, yeah, finish out the Pearson thing and I’ll go back in a second.

[00:15:52.670] – Kerry
Yeah. So a Pearson’s correlation, it’s a number that’s between negative one and 10 is there’s no correlation between the two sets of variables. Negative one means there’s a perfect one to one linear correlation that’s negative, and positive one means there’s a perfect linear correlation that’s positive. If you score up 0.3, that’s considered. There is a relationship of some kind. It’s weak, probably not linear zero six and above, like the absolute value of zero, six and above, that’s a strong correlation. And the thing about humans is humans are really messy and they have a lot of chaos in their hearts. And so you don’t actually really ever get very strong correlations with humans. And so I wasn’t really expecting to get anything that was above a 0.6. And what I actually did find is once so we tested each of these hundreds of factors one at a time to look for any correlation. We had the eight that did have a correlation to sales, and the number one correlation to sales was a coefficient of 0.4, which was pretty strong. Not pretty strong. It’s not technically strong. It’s technically in the weak moderate area. But humans again, are super messy.

[00:17:08.650] – Kerry
So that’s the size of the author advance. The next one was the relative fame of the author. And we included in that if they’d won any major awards. So like if they had won a newberry, they got a higher rating than if they hadn’t. For example, the third one was carryover. And so that was an association with something famous. In my dissertation, I called this The Twilight Effect because there was a little while where if you had a book about a vampire, it would sell super, super great. And it would do that because Twilight was selling well. And so the book didn’t actually wasn’t Twilight, it just was associated with something that was famous. So I called that carryover. It’s that association with something else famous.

[00:17:50.530] – Scott
Yeah. So that Carryover is that carryover is almost writing to trend, kind of.

[00:17:57.000] – Kerry
But the problem, like you can’t predict.

[00:17:58.880] – Scott
What’S happening to be on trend.

[00:18:00.740] – Kerry
Yeah, you happen to be on trend, right. Happened to be on trend. And not in the bad way. Like writing about a plague in 2019, that was not quite the right kind of psychic. You need to be like the right kind of psychic if you want to write on trends.

[00:18:17.050] – Scott
Right.

[00:18:17.250] – Kerry
So that one, it’s not really in your control. There’s some things that you can do to increase your carryover score. You’ll see people do new adaptations of fairy tales or something and that will give you a little bit of an advantage, but not as much as writing about vampires when Twilight happens to be going crazy with sales. And so that was the third strongest of the eight that I found. The next one was a cover likability score. I did a cover likability study three different times with two different groups. One, I did a high school group. And one thing that was interesting was the high school group did not show any correlation at all to sales. And I thought that was OD. And then I did it with college students and it was almost as strong as fame. The COVID like ability correlation with sales was, and I think what it is, is that high school students are less likely to buy their own books than college students are. And so if college students liked the COVID it was more likely to sell. That’s my theory. I can’t verify it, but that’s what I think was happening.

[00:19:15.810] – Kerry
So that was up to 0.4, which is about as strong as fame. The number five one was number of Amazon reviews, which I dropped off of the bigger calculation because I was pretty sure that was not causative number of authors, Twitter followers came in as one of the factors, which was interesting. I started tweeting after that. I was like, okay, fine, I can do that one. I can tweet a lot.

[00:19:40.410] – Sunyi
Not anymore.

[00:19:41.710] – Kerry
Well, not anymore because now Twitter is dead. And twitter was particularly effective at being correlated to book sales. And I think it’s because twitter is a reading medium. And this is again, just speculation, I don’t know for sure. And people that are on Twitter are people that like to read. You don’t go to Instagram because you like to read. You go to Instagram, you like to look at pictures, and like half the time you don’t even read the little captions. But on Twitter there’s only the captions, and that’s what people do. And so, yeah, that was one of the factors, was Twitter followers. And then the 7th thing was starred reviews in particular locations. So for Ya, the school library journal was really important. Publishers weekly and book lists were really important. And the last thing that was correlated with sales was crossover appeal. So I was studying Ya fiction, and if it had the ability for adults to buy it, it did better. So those were the eight factors that were correlated with sales. And so those were the only factors that were correlated with sales. Out of something like 300 different factors that we tested, those were the eight that had a correlation with sales.

[00:20:46.340] – Kerry
They are all marketing factors and almost all not in the control of the author. And so then I created an algorithm using those and estimating sort of the various sizes of those correlation to look for a correlation with sales in general. And so now I have an algorithm that can predict book sales. The Pearson coefficient is 0.7 of marketing and sales in the book market, which was kind of an astounding number. When I got it, I was like, humans are not that neat, but this is a neat number. This is strong, this is a strong correlation. Like it’s very strong. Zero, seven. Like I didn’t expect to have anything higher than that.

[00:21:26.490] – Scott
So yeah, can you explain how that 0.7 came about? Being that these individual factors were all in the 0.4.3 range, what does that 0.7 represent? Is that the aggregate of all of these? What’s the difference in the analysis there?

[00:21:47.880] – Kerry
So it’s like an aggregate. You create an algorithmic sort of equation that looks really impressive when you write it down on paper. It’s not really that impressive in real life. Basically what you sort of do is you add all of those things together and multiply them by the strength of their original correlation, and then you have this new additive thing that you can make the prediction that way, and you have a new set of numbers and then you can run the correlation. So basically what we could do is for any individual book, if I knew how much was paid in the advance and I could figure out the rest of the advance was the hardest part. And that’s why we use the publishers marketplace to do that, which was great because authors love oversharing, and so there was plenty of data to draw from there. But once you have all those things, you can plug them into the algorithm that we made, and it will come out with how likely it is to sell or not sell based on those numbers. And it’s accurate about 75% of the time, actually. So we can predict at the point of acquisition with about 75% accuracy, how well a book is going to sell.

[00:22:56.280] – Sunyi
So if we gave you numbers, hypothetically, can you plug them in?

[00:22:59.660] – Kerry
I can, and I have done that.

[00:23:03.210] – Sunyi
Yeah, because we had our first year s,ales and we have a rough sense of it. Sorry, I know you’ve got a bit of lag, so we’re cutting into each other.

[00:23:08.990] – Kerry
Oh, sorry. We’re literally all around the globe right now. It’s kind of awesome. Yeah. So if I had the numbers, I could plug them into the algorithm and predict whether it was going to sell or not. And I have done this when I’ve consulted with publishers before, and one of the most important things if a publisher will ask me why is this book selling or not selling, is to figure out if it’s hit the marketing viability threshold. Because this was something that I started to notice when I put all of the final numbers together in the sample of these 500 books that I had. I expected there to be like my initial idea for the dissertation was I wanted to look at the outliers, like, I wanted to look at the books that didn’t get marketed, that sold well anyway, and I wanted to look at the books that got well marketed that flopped. And what I found was there were not really any so when it came to well marketed books, there were definitely flops. There was plenty of books that actually really tanked. I would actually say a lot of the well marketed books tank.

[00:24:11.850] – Kerry
If you’re looking at how well they should sell versus how well they do sell, probably four out of five of them do not sell as well as they should when they’re really marketed. But when they do sell, they sell like crazy. They spike off of the charts. Yeah. From the author perspective, if you earn out your advance, you’re probably going to be fine, and they’re not going to look for an excuse to punish you, but if you don’t, that’s where it can get dicey and tricky. But when it came to the lower end of the market, I found that I was looking for any sales spikes, like at all. I wanted one book. There was literally zero. There was literally zero for a huge chunk of the sample that I had. And what I found was until you hit this marketing viability threshold, you had zero chance at breaking out. Like, not even not even like one in 100. It was literally there were zero books that did it in the sample. And so often, if I’m consulting with a publisher and I run all the numbers, I can say, this didn’t hit the marketing viability threshold.

[00:25:17.980] – Kerry
That’s why it didn’t sell, period. It gets a little trickier if it’s a little higher because there is some chaos in it, because I think marketing is a little bit like virality. Like if you’re on Twitter, rest in peace, Twitter. But when you’re there, you never know what’s going to go viral or not. A good tweet, maybe it’s going to go viral, maybe it’s not. But the quality of the tweet has nothing to do with whether or not it’s going to go viral. It might be more likely to go viral, but a good tweet is not enough for it to go viral. There has to be other things that are at play, and I think that happens with book marketing too, where a good book is not enough to sell, and even like a good book. Plus marketing is not always enough to sell, and what actually makes it sell is a little more chaotic. But you will have zero chance at selling if you don’t hit the marketing viability threshold.

[00:26:06.570] – Scott
Just to review, did you define that marketing viability threshold? Do you know what comprises that? Or rather, how did you define it?

[00:26:16.550] – Kerry
One of my publications is actually on this. This is called marketing and sales in the US. Young adult fiction market. It was published in New Writing, which is the International Journal for the Practice and Theory of Creative Writing. And so what I did in that article was sort of put those numbers out there. And basically the way that I had gotten the total marketing score for any book is I took each one of those individual factors and gave them like a score. So if you had a really high advance, like, you might get five points instead of one point, and if you have no really low advance, you’d get zero points for that. And we would add all those points up in the end and then see how high it got. And if I remember correctly, in that sample, it was five. I’d have to actually look it up and see what the number is. But you had to hit a certain number in that score before it got up. And so usually what that meant was you had to get an advance of you had to do sort of one of a couple of things.

[00:27:17.900] – Kerry
Either one had to be famous already, or two, you had to win a major award. Or three, you had to have a really likable cover. And when it came to things that you could have control of before publication, you had to have a combination of things. Usually, like, you had to have a likable cover and you had to get in advance of more than $50,000, or you had to get like, maybe you didn’t get in advance of more than $50,000. But you got a three book deal and had a likable cover and you’ve got a lot of Twitter followers that would get you over the marketing threshold or not. And so it really just depended. It was a book by book thing. And so I’ve got this massive spreadsheet that has thousands of pages on it and I can plug numbers into it if I know them for an individual book. And then I can tell you like, oh, you scored a four and that’s below the marketing threshold, or viability threshold or not, or I think in the end, the highest score I think I ever saw for a book was Twilight. I ran the numbers on Twilight and it had like something like a marketing score of 16, which was the highest that I’ve ever recorded.

[00:28:18.080] – Kerry
Most marketing scores were like one or two when we did it, they were really small, they weren’t huge.

[00:28:26.160] – Sunyi
But that’s funny. I think the only thing I had going for my book probably was the advance in the marketing. And that was something we were talking about in discord because it did sell well this past year. And I know that a lot of books don’t, and I was using the example, I don’t have your brilliant technical language, but when I explain publishing to people who aren’t in the industry, I describe it to them as like trying to build a rocket where you need a certain amount of fuel to get lift off. And the goal is to have a rocket that gets in self sustaining orbit. And something like Brandon Sanderson. I think of Sanderson as like this giant moon, right? He’s in the sky, he’s not coming down. If he never wrote another word for the rest of his life, he would still be this icon floating above us as a satellite. He would be fine. Same for J. K. Rowling. She could literally set everything she owns on fire and never write another word. She would still have enough money, stephen King as well, et cetera. But most rockets, it’s like they just don’t get the lift off, so they don’t get the fuel, they don’t get the ground support.

[00:29:23.520] – Sunyi
They fly a bit and then they nosedive. And that’s kind of what most book launches, is what they look like. You’re there for a month and then you’re off the shelves and you see your sales tank and it’s so difficult to get to that threshold. I did wonder if I know it’s not really your area, but obviously indies are having to do all that stuff themselves. I don’t know if you ever looked at them at all.

[00:29:47.610] – Kerry
And yeah, I didn’t study the indie market, but I want to at some point because I have a suspicion. One of the things I’ve always wondered a little bit that I would like to have a chance to test someday, I would need more research assistance to read 300 books because I don’t want to read 500 books again. I did that actually during my dissertation. I spent an entire year doing nothing but reading Ya books to get these 200 books read. So I read one book a day, basically for an entire year. It was a rough year, it turns out, because a random sample, we’re better at choosing books that we were going to like than we think we are. And when you have a random sample, you end up reading books that were not meant for you. You’re not the audience, and it can be very painful. But yeah, one of the things I do want to test someday if I have enough research assistance to do this or get another grant big enough to pay them to do this, is I would like to repeat this in an indie market because I have a suspicion that quality might matter more in the indie market.

[00:30:44.050] – Kerry
Mostly just because there’s already gatekeeping and publishing. You’ve already made it through a slush pile. You’ve already gotten an agent, you’ve already gotten quired from an agent in submissions. There’s several levels of sort of quality check before you get to the final stage. Even so, some books get published and we read them and it’s like, I don’t know why this got published. And then other books are beautiful and get published and just don’t get published, even though they were beautiful books. I’ve read a couple of just wonderful, amazing books that never found a home, and I was always sad about that. And I actually found a couple of gems too, when I was doing my dissertation where the books were just so good and they didn’t get marketing. And I was one time one of the books I was so sad about. I literally called the editor up on the phone and I was like, listen, this book is amazing and this cover is killing it. We need to redo the COVID And she blessed her. She was like, okay. And she listened to me and I sent her pictures. I’m like, here’s what you need for a cover in this market.

[00:31:42.430] – Kerry
And they redid the COVID and it started doing better. And I was like, okay, yeah. But I do have a suspicion that maybe it might matter in indie books, but I’m not sure because I don’t have the numbers to back it up.

[00:31:56.850] – Sunyi
I almost wonder if there’s just so many indie books that it’s like statistically I don’t mean that as a bad thing, but just there are statistically so many indie books that something is bound to succeed. Almost a kind of spaghetti at the wall approach. But then that’s the same with Trad, when books that get marketing, like big, big books that fail, is that like relative to their advance, if you see what I mean? Or is it still like, yeah, failure looks different at mid list from lead title, if that makes sense.

[00:32:24.370] – Kerry
I wasn’t working with the actual number of sold copies of books because of the tobacco with not Pearson Nielsen. That wouldn’t give me Nielsen. But my sense was that sometimes really male marketed books fail badly like they do just as badly as the non marketed books, but they’re also much more likely to spike. And most of the time you get in, what you put out tends to be what I found in the data. Like whatever they put in in marketing, you would usually get almost that back. Not always, but usually you would at least break even. And so publishing is sort of this weird game where any individual book isn’t really making them any money, but they thrive off of those breakouts, the ones that sell millions of copies for no reason that they can identify in the courtroom. And those are where you get a lot of the profit that happens in publishing and creating that sort of thing. It requires a great deal of marketing, but marketing isn’t a guarantee for it. And so it’s like one of those.

[00:33:32.120] – Sunyi
Tricky, tricky, very aware I’m sorry, I’ll say this last thing. Let Scott take over. I was very aware over the past year that the things that made the biggest difference to my book sales were almost like these tiny individual decisions by specific people. The Crate team liking it, or the right person at BNN saying, oh, yeah, we’ll take it for a monthly pick. Or the right person at Waterstones agreeing to do a special edition. And there are about five or six decisions, I think, by basically a handful of people at the right time and right place that made a difference of thousands of copies. Literally. If you get into like a monthly pick, it makes a huge difference. If you get into all these other things, special editions or crates and all that, they all really help, or the right review at the right time. And it’s just it’s staggering that it’s like you could have a near miss of that and it will be a difference of 10,000 80,000 sales or something. Sorry, Scott gone.

[00:34:26.580] – Kerry
It’s true. It’s true. And some of the things that ended up mattering a lot, that were strange were things like the book just had a really prominent display at Walmart because somebody at Walmart decided to make a display of the book. And it’s like one person’s decision will change the life of that book forever. It’s interesting.

[00:34:45.850] – Scott
As I’m sitting here listening to all of this, reading through your excellent PowerPoint that I believe we will link in the show notes.

[00:34:52.820] – Sunyi
Absolutely.

[00:34:53.870] – Scott
Your top factor is size of author advance and all of these things that both of you just mentioned. So Sunyi with her Book of the Month pick, somebody at Walmart making a display book, making it on a display at BNN, whatever it is, I suspect even Editor Picks on Amazon, et cetera, those are not so random as they appear, right? And that’s something that SUNY and I and our discord and various people talk about quite a lot is that a lot of those things happen because behind the scenes, the publisher puts an effort into making that happen, right? As we learned on our bookseller episode, publishers have sales reps that go talk to bookstores and sell into bookstores, basically tell them to order a certain amount. They have people who are pushing these things in all of their various channels where they can sell. And my theory is I’m fairly positive that those activities behind the scenes, even simple things like whether they print arcs for a book, right, those are all highly correlated to the size of the author advance because they have to make their money back, right? Like, if they handed out a lot of money, they then assign internal resources to go get that money back, at the very least.

[00:36:24.840] – Scott
If not get some multiple back to make it worth their time from a profit perspective.

[00:36:31.690] – Kerry
And they’re typically making those decisions ahead of time, too. Like the editors are filling out profit and loss estimations when they’re going to acquire a book where they’re like, this is how many we think we can sell, and they compare it to other books that maybe have sell, but it’s all just like a guess. And then it sort of becomes a little bit of a self fulfilling prophecy because if they think it can sell a lot, they give it more marketing and then it does. And I think one thing that’s kind of interesting, I think about all those little points you were talking about on the sales distribution channel is some of the decisions about a book. They’re not necessarily made by people who have read those books. Sometimes they are, which is great when they are, but they’re not always. Sometimes it’s just like, oh, we need a Ya book or something, and this is the lead title that gets pushed a lot and it’s got a pretty cover or something like that is enough to make a book sell sometimes.

[00:37:19.490] – Scott
Yeah. And to your point, none of these things that ended up as high correlation items with sales are things that would necessarily influence an acquisition and a PNL at acquisition, right? Like maybe the relative fame of the author if it’s an existing author with other works. Right, that certainly matters. Carryover or writing on trend might matter, but publishing moves so slow that even an editor who’s acquiring and owns the schedule for their publisher doesn’t necessarily know what will be breaking out at the time that that book’s coming out, right, because it’s not necessarily up to just them. It’s a broader market trend that they don’t control. Cover that’s not determined at time of acquisition necessarily. Number of Amazon reviews, obviously not author Twitter followers, maybe. But yeah, I mean, just the degree to which things are determined right at the point of acquisition is pretty crazy.

[00:38:26.320] – Sunyi
And this is like, I don’t have proof for this, but this is my private theory that I think this is why high concept has become such a dominating phrase in publishing. Because when you have a book that’s high concept, I assume most people know what that is. But in case any listeners don’t, high concept basically is like a book that’s easy to summarize, that sounds good as a pitch. So you can summarize divergent as, like, not divergence hunger Games as kids are fighting in gladiator battles for food, whereas Old Man of the Sea is not high concept because it’s a man goes fishing, and it’s all in the execution, and the concept sounds a bit pounced, so you actually read it. But high concept is easier to sell in that way that if you have a book you’re trying to get what Nick Binge called the Hype train, it’s a lot easier for your editor to go to the sales rep and say, here’s this snazzy idea. And for the sales reps to go to their BNN stores and stuff and go, here’s this snazzy idea. And all the way down the chain, if the book is easier to sell, that’s helping it as well, I suspect.

[00:39:23.170] – Kerry
I think that’s absolutely true, and I think it’s like a Tweet length synopsis kind of is what you have, really. And I think part of that is that the way the successor of the book is determined at acquisition is through the size of the author advance and whether or not it’s going to be in a series and how many books are required and whether or not there’s an auction. And part of that is in the room of the acquisition meeting are the marketing people, and they’re not the people that have read the books, and they’re going to decide, how many books can I sell based on that pitch? That very short synopsis. And so that Tweet length pitch becomes really important for a book’s success or failures, like the ability to sell it in like, 200 characters. And if you can, it’ll do better because you’ll convince the marketing people, and if you convince the marketing people, you get more money. And that’s what sells books, is marketing.

[00:40:19.180] – Sunyi
Yeah, I used to wonder why everyone kept summarizing my book as like, oh, it’s just about eating books. I was thinking, that is like 2% of the actual novel. And then I realized that that was the hook that they were selling it on, if that makes sense. People are reading it thinking, oh, it’ll be about people who eat books, even though it’s not really and I think that disappoints some readers. But anyway, that’s a whole other thing.

[00:40:40.050] – Kerry
No, it’s true that short pitch matters a lot. It’s what creates that sort of viral earworm that gets people to try it, and then the book itself can do all sorts of things. And if it’s a surprising book, like, great, awesome. And then if the book is exactly what you expected it to be. That’s also good. So it just is really important to have that initial short marketable pitch.

[00:41:04.910] – Scott
And just so I hit on this to clarify, you did look at, even if they were subjective, you looked at quality of writing type categories, right? Factors. How did that go? Like, how did you measure subjective quality factors and what did the analysis look like?

[00:41:26.460] – Kerry
So I did a couple of different things to try to get at that kind of a question. And so on a very simple level, one thing I would do is I would rate the quality of book, like scale of one to ten, and then I would do paragraph level ratings, like what’s the sentence level, paragraph level of style? Like how complex is the language? And I would do that, or I would do I tested something like twelve different emotional words. Like, this book made me feel stressed, this book made me feel romantic, this book made me feel anxious. Or like a bunch of, I don’t remember exactly how many, I think it was ten or 15 of those yearning was the one that actually had possibly a result that was happening. And then I also used three or four or five potentially different plot structures. So we looked at different plot theories of how stories are shaped and formed. We tested the Campbell’s Hero Thousand Faces Monomith plot. It came out as being way too, like, sexist and racist, which shocking, everybody that knows Campbell is shocked by that and was not terribly predictive of how a book was going to do.

[00:42:41.440] – Kerry
There was one plot theory, though. Somebody had written a plot theory that was supposed to be a female counter to the Campbell cycle because Campbell is like super masculine, and so somebody did a feminine version of it. And that one actually did occasionally hit some points. Not until we factored marketing out did it start to get close to it. But I think that was a demographic thing that happened. But there were some interesting disconnects that I found when I started to compare some of those different metrics against what was selling or not selling or what kinds of marketing a book was getting or not getting, which was like that. Books that were too female were not getting as much marketing as they should have, but most of the reading market was female, and so they were selling much better than they should have because they just didn’t get the same marketing as another book did. Same thing was true with race, where books by it wasn’t just books by people of color, because those were pretty consistent number wise. But when you started to factor in a couple of different racial factors, you found that books that were about race, books that talked about race, they did not get as much marketing as the other books did, and so they had correlated low sales.

[00:43:54.190] – Kerry
But when you did the I called it a residence score. When I subtracted how well a book did sell from how well it was supposed to sell based on the marketing, like, the marketing would predict a certain number of sales. And then I’d compare that to how it actually sold. And those books were selling really much better than their marketing predicted that they would. And so that was an interesting sort of disconnect that was in the data, too. I think it was a confirmation of something that a lot of people already knew about the world at large. But it was interesting to see. There was a lot of different sort of things that we tested that were like that. But most of the ratings myself so I guess there’s some bias in what did I think was quality or not, but nothing that I could as a person with a PhD in writing, could identify as quality had any correlation to sales.

[00:44:48.670] – Scott
One of your categories is number of Amazon ratings. Did you look at actual Amazon scores or goodread scores and how those correlated?

[00:44:58.670] – Kerry
Yes. So we recorded I don’t think we did goodreads because it wasn’t as robust back in I was working on this in 2011, was when I graduated. But we did Amazon reviews. We collected all of those. And there was a correlation between the Amazon not the score itself, not how many people liked it or didn’t like it. There was not a correlation between that, but there was a correlation between the number of reviews. So better selling books got more reviews. And I don’t feel like that’s a causative correlation. And so when I created the algorithm to predict how well a book should sell based on marketing, I didn’t include Amazon ratings when I did that.

[00:45:39.900] – Scott
Sure. Yeah. So the number of reviews mattered, but the actual score, like a three or a four or a 4.5, et cetera, did not very interesting.

[00:45:49.570] – Kerry
Yeah.

[00:45:50.590] – Sunyi
Scott did a very, very small, informal study where he looked at that, and that was something he was very convinced of as own good reason. On Amazon, it kind of doesn’t matter if your book gets like three stars or 4.8 stars.

[00:46:02.470] – Scott
Well, yeah.

[00:46:05.410] – Kerry
If people hate a book, it can be good.

[00:46:07.450] – Scott
I was really bummed as I, you know yeah, I was really bummed as I watched my goodread score sink over time, which is something that happens to all books, even ones that end up with a really good score. And I spent a good amount of time thinking, well, I guess I just didn’t write a good book. Time to move on, right? But one day, I just decided to compile the scores of a whole bunch of books that had released in the last two or three years. And there are some really weird anomalies of books that end up with really high scores that it just doesn’t make sense to me, at least. Like, just weirdly high like these books that don’t seem any more popular than others or any better than others in any obviously subjective way end up with like a 4.5 and stick there. But anyway, I looked at, I think, close to 50 50 or so releases from either debut or newer authors in the last two or three years, and there was absolutely no correlation between goodreads I didn’t look at Amazon scores, but goodreads scores and the number of goodreads scores, if that makes sense.

[00:47:18.670] – Scott
Which number of goodreads ratings was my proxy for sales? And I was very surprised to find that I had believed at least a little bit more in goodreads ratings before that.

[00:47:33.970] – Kerry
But yeah, same was true with Amazon. We couldn’t find any relationship between the actual score or the number of stars. Just the number of reviews was the only correlation. Like basically any rating of quality that we could think of, that we looked at. There was nothing that connected quality to sales. I still haven’t found anything. If you find something I’m looking for, I would like to believe it’s there. I don’t know that it’s true. Having been on Twitter for a couple of years now and seeing how random it is, what goes viral and what doesn’t, and when is more likely to go viral and when is not, and to kind of understand that, algorithm there. I think I understand a little better how the marketing does work and that it won’t necessarily ever have anything to do with quality alone. But I think every now and then, quality, I still have this hope that maybe like a really great book that takes your breath away, maybe that will affect it. And I think that it’s true, it does. But it’s just not enough to create a big spike in sales. It’s never going to be enough to actually drive the sales.

[00:48:39.560] – Kerry
The sales is going to be driven by outside market forces because publishing is a business. It follows laws of business, not laws of art. We have these weird little capitalist brains that have tied those two together that says that beautiful quality art is going to sell really well because we have this prosperity gospel that we believe, but it’s just not true. The books that are approached the way a businessman is going to approach it and that sell it the way that people sell things like those are the books that are going to do well and it doesn’t actually have anything to do with what’s inside the book.

[00:49:14.180] – Scott
Yeah, and I think it’d be too hard to measure anyway, right, just because taste is so subjective and even quality is so subjective, right, like writing that I find horrible and unreadable. Other people don’t seem to mind at all and vice versa. But I think it’s worth pointing out that something you said earlier, that even the strong correlations per your study are statistically moderate or even weak. And so there is a lot of chaos still built into the system. It’s just that you mentioned your 0.7 Pearson coefficient for marketing as a catch all category, and particularly in terms of hitting that minimum marketable or minimum marketing viability threshold. I think that makes a ton of sense. I think there’s still a lot of subjectivity and quality, quote unquote, that goes into it. But yeah, without having hit that threshold, people just don’t even know about it to give it a chance to grow on its own.

[00:50:21.400] – Kerry
That’s true. And I think that’s really the core of it is people can’t read a book until they’ve heard about the book. And hearing about a book, we want to believe in word of mouth, and it’s possible that that has some sort of effect, but it doesn’t have a measurable effect that I could find a way to measure. But the marketing is how you get people to hear of a book, and if you hear of the book, then the book sells.

[00:50:44.980] – Sunyi
There are some examples I can think of which are outliers, and people often cite them in these conversations. I guess the classic one that I hate having dragged up is 50 Shades of Gray, which was like, apparently a genuine world amount of success. But I don’t know how you feel about it. But my sense on it is that these outliers are just so far outside that we can’t really learn anything aside from the fact that outliers exist, if that makes sense. You can’t replicate 50 Shades of Gray, you can’t replicate JK.

[00:51:10.620] – Kerry
Rowling, you can’t replicate it, but you probably could have predicted that it would have done well because for one, it would have had a high carryover because it started as Twilight, so it had an existing group of people that already were into it. So that already puts it that high carryover score, probably gets it above the viability threshold. And it also like there’s a couple of other things that it particularly had, like it had popularity before it was picked up was another thing, which means that then you’re going to have a higher advance. And I don’t know the actual advance for that book, but I imagine that it was probably even if it was just like one category above the lowest level because there’s the nice and the very nice. It only took like one category above the normal level. Most books, like 80% of books were only like, got in the nice category, which was up to do you remember if it’s up to $20,000? Up to $50,000 on Publishers Marketplace?

[00:52:06.790] – Sunyi
Well, 49 nine, which is like up.

[00:52:08.750] – Kerry
To 50, yeah, 49, yeah, sorry, I.

[00:52:14.200] – Sunyi
Was just going to say I think she’s picked up by a small press first, but then an editor, allegedly, I mean, these stories always spun into myths.

[00:52:21.800] – Kerry
But they are it’s part of the marketing.

[00:52:24.310] – Sunyi
Allegedly an editor heard people talking about it at the school get yeah, the marketing for JK. Rowley. She’s rejected.

[00:52:31.130] – Kerry
Not really. And she actually got a lot of money for the books too. And Stephanie Meyer got a lot of money for those books. It was something like a $600,000 advance that she got for a three book deal. And so that was enough to put it in the major deal category, the highest category. So even if you’re just in that very nice category that you’re making a little bit more than $50,000 in the advance, you only need a couple of points to get into the viability threshold. It’s just that most books don’t even get that high. So like the vast majority of books, I think like 80% of the sample was in the nice category, which was less than $50,000.

[00:53:06.590] – Sunyi
In one of our episodes, we did a kind of direct launch day comparison for me and Scott and basically found that it was correlated quite strongly, where I had ten times the advance he did, and I had about ten times the sales on the UK side. Sorry, USA side can’t compare UK because he didn’t get a UK release.

[00:53:26.530] – Kerry
Sorry.

[00:53:31.410] – Scott
I kind of hate that. This carryover category that you’ve mentioned and been talking about a little bit is somewhat justifying the retelling trends that we’ve seen lately, where everything is marketed as.

[00:53:42.900] – Kerry
A retelling, even if it’s I mean, it’s a decision. That’s a marketing decision. It’s not an art decision. It doesn’t lead to better art, it leads to more sales. And the people that are making those decisions about green lighting it are the marketing people, is why. So the retellings, that’s why there’s a lot of them and that’s why there’s a lot of remaking of movies, is because it will sell. They know it will sell. It’s more of a guaranteed deal.

[00:54:06.590] – Scott
Yeah, totally makes sense. So just a couple of questions that I picked out from our conversation. So one of your categories is Twitter followers, and you I think, mentioned Instagram briefly, but did you look at other social media sites and other places where authors might have followings and analyze those or was Twitter the only one?

[00:54:28.220] – Kerry
We recently looked at TikTok because we’d heard a lot about book talk and everyone’s like, this is how you’re going to sell a lot of copies. We were not able to determine a correlation to TikTok number of followers or number of videos or number of views for a particular video. That could be partly because when you have a random sample, a random sample has drawbacks and it has advantages. Like the advantages you get a clearer sense of what’s actually happening without the spin, without the people saying like, oh, this was such a surprise, when it’s really not a surprise at all mathematically when you look at the numbers. And so the random sample is helpful that way. But when you’re trying to go from the other side and you don’t have that random sample, you sort of just see the. Product of all of that and you don’t actually see what went into it. And so one of the things that was tricky about looking at TikTok was that most of the authors in our random sample were not even on TikTok. And so there was a very small percentage of the authors who had any sort of TikTok presence.

[00:55:29.200] – Kerry
And it was not enough to determine with any sort of statistical significance whether or not it had an effect. My instinct would be that the bigger of a following or audience you get, the more books you’re just going to sell. Because any sort of fame is fame and fame is correlated to sales. So if you’re TikTok famous, you’re probably going to sell more. But that’s not something that I was able to reproduce with numbers on TikTok.

[00:55:52.890] – Scott
Yeah, that’s very interesting. And being that this was put together back in 2011, I wonder if there was some microcosm of Twitter at the time and whether people were more engaged versus now where there are a lot of writers on Twitter and maybe not as many readers.

[00:56:11.170] – Kerry
Yeah, I mean, it’s hard to know. Although every time I get a grant for more research assistance, I update my sample and we expand it a little bit more and we gather more data. And so far Twitter has always been significant. Like it’s always had an effect on sales, the number of followers, and at a lower number than you would think. That was partly I think I decided to finally try to write another book in 2019, was that you only needed to have 3000 followers before you started to notice a mathematical effect on the sales. And then after that, every 5000 more it would go up and up and up. And so by the time you have, like I have almost 25,000 followers right now, that alone is enough to put a book into the viability threshold. 25,000 followers on Twitter is. So that was something that stayed consistent. And yes, the demise of Twitter has made me sad as somebody that studies book marketing because there is not a good alternative yet, especially not one that’s big enough with enough effect that we can measure yet. So it’ll be years before we know if anything else is going to replace Twitter with that.

[00:57:21.420] – Sunyi
I’ve got two questions, if that’s okay. So the first is, did you look at Twitter followers in relation to ratios? So, like, people who have 20,000 followers but they follow 20,000 people. And the second question is, what is publishing’s response to your data? I mean, you mentioned teaching publishers why their books didn’t sell, but does anyone actually act on this or take it on board?

[00:57:43.590] – Kerry
So yes, we looked at both the ratio of followers and following and it didn’t really seem to matter. I think part of that was that most people that have a lot of followers aren’t necessarily following everybody back. And so there was only a couple of people that followed 100,000 people and then had 100,000 followers but it did not affect the numbers as a whole. So when we aggregated it together what we didn’t do is I didn’t take all the people at those kinds of artificially inflated ratios and test them alone and separate them but as part of the group it didn’t seem to matter. So just the number of followers was the one that there was the correlation. Nothing else seemed to have any sort of correlation. And then the next question you said is like what do publishers do? They don’t usually tell me what they do when I talk to them. Every now and then I’ll consult with them and I’ll talk to them about some books or things and sometimes I’ll see the book rereleased with a new cover or something or they’ll tell me like oh yeah but they never are surprised by the data.

[00:58:59.530] – Kerry
I think that the editors aren’t exposed so much to the numbers and the hard marketing things that the marketing department is. So sometimes they still have that secret hope that they can find a book that’s really great and it will sell because it’s great, not because it’s marketed I think but they’re never terribly surprised to realize like oh no, it’s actually true but it’s just the marketing. I mean I think one of the limitations is that there’s only so much money you have ahead of time to invest and so if you’ve got this big pool of money and you’re going to vested it’s not unlimited. And so they have to decide where are they going to put it and so they tend to try to put it behind the books that they already are thinking are going to do well so they don’t lose a bunch of money, because they can lose a lot of money on a book like it’s happened before, and it happens frequently. I would say actually more books don’t quite sell as well as they would hope than the ones that do and there’s always this sense of surprise when a book does a lot better than expected.

[01:00:02.260] – Kerry
Even the people in the publishing are like oh hey, look what we have here, this is exciting. Most of the people that I threw.

[01:00:11.110] – Sunyi
It off a cliff and it flew.

[01:00:14.850] – Kerry
Most of the publishers I met, they want the books to sell. And not just that, they’re not just like craving they don’t go into publishing because they just want to make a lot of money. That’s kind of funny, actually. But they are constrained by the fact that it’s a business. And if their books don’t make money, then they don’t have a business. And I think that’s hard for a lot of them because they have these idealistic senses of like I am going to be the one that makes the quality book sell. And I think that they can influence things, they can influence trends and they can choose things or not choose things. But some of it is kind of out of the immediate control of the editors. For sure. The marketing department, not as much. But even then, they’re looking for very particular things in the marketing department. They’re looking for carryover, they’re looking for something they can pitch. They’re looking for something easy to blurb, something easy to hook and sell in a very short amount of time. And that’s not necessarily something that has anything to do with the quality of the book, because some books are very nuanced and complicated and you can’t give like a tweet length summary of the book and have it sound like anything good.

[01:01:31.790] – Kerry
That ability to do it quickly is a marketing key that’s important.

[01:01:36.380] – Scott
Yeah. Per your analysis, it still just makes no sense to acquire a book, develop it, publish it, and not reach that marketing viability threshold, though. So I hope that that’s something that publishers catch on to.

[01:01:53.730] – Sunyi
Yeah. On a basic level, I don’t think they’re wrong to do that. And I think we are all grown ups. We understand that some books have a wider commercial reach than others. We understand that Stephanie Meyer books are just going to sell more on release day than, like, random person John Smith in a corner. And that’s kind of fine. It’s just I think where my issue comes into it with is that we publishing seems very invested in believing in this myth, like this cultural industry myth, that all the books are starting this foot race on an equal footing, when in fact, some of them are starting this foot race without shoes, while others are literally, like, on a motorcycle or something. And that’s the part that bothers me. I know that not all books can get equal funding. It’s the pretense that we are all on the same footing, same competition with the same yeah.

[01:02:42.340] – Kerry
And I think part of that is like a marketing thing itself. Like, you want to have the myth of a book, the myth of the hidden gem that came out of nowhere. It’s like the author who wrote her first book when she was three. And these are stories that are told. Marketing is a story that’s told. It’s not actually has anything to do with the reality. I think that the publishers, I think they don’t always know what the minimum marketing threshold is as part of the problem. They always are a little surprised every now and then when I find a book, they’re like, we don’t know why this book didn’t sell. It was amazing. And it’s like, well, it got a score of four that didn’t hit the viability threshold. And they’re like, what? What do you mean it didn’t hit the viability threshold? So they don’t always have that same very clear mathematical algorithm in front of them the way that I did. They don’t have the thousands of spreadsheets of data. I mean, I think somebody at the publisher does, but it’s not the editor who’s doing the profit and loss sheet together that’s trying to sell a book to the acquisitions committee.

[01:03:41.270] – Kerry
Yeah, it would be somebody else. And I’m not familiar enough with the inner workings of the publisher to know exactly who that would be. Was it the Random House publisher that said in court that they just had no idea what made a book sell or not sell? I don’t think that’s wrong. I don’t think that they always do have a sense like they do, and they don’t want to. I think not a single one of us wants to give up that secret hope that quality is enough to make a book sell better. It’s just not. And that’s a sad thing you have to accept if you’re going to write books, is to know that writing is one thing and art is one thing, but publishing is a business. It’s not going to follow art rules. It’s going to follow business rules. And that’s a hard thing for artsy people to swallow. I think even publishers, it’s hard for them to believe that that it’s true. Even if they’ve seen the numbers and they’ve seen it happen over and over again and they’re depressed by it themselves. They don’t know how to change it, or they just don’t want to accept it.

[01:04:40.510] – Kerry
I think I feel that I secretly think that my next book is going to be amazing. It’s such a good book. It’s going to sell so well. Yeah. And we’ll just see how that goes.

[01:04:52.140] – Sunyi
But I just can’t believe they’re not beating down your door desperate to get this algorithm. I mean, if I was an editor, I would be scratching at the window.

[01:05:03.170] – Scott
These publishers have much more granular data than you are able to access. Right. As mentioned before, that author advanced size is really a proxy for many different things that happen behind the scenes at publishers. And so my frustration is that there is apparently nobody doing basic linear regression of activities put into a book, and the result like, Just fucking do it. It’s not that hard. You can do it in Excel. You can do it like, just do it. Linear regression is something that I think I learned in 9th grade. It’s not that hard anyway. I have to get at least one rant in. I have to get at least one rant in per episode. And I think that’s what frustrates me the most is that they pretend that it’s like, oh, well, you didn’t have the platform, or oh, your book just didn’t catch on with people. And it’s like, well, okay, but how many people saw it and what did you do and what factors matter? And it really was frustrating to ask what factors matter and just get to get nothing back.

[01:06:16.060] – Kerry
I think it’s a combination of the fact that I think some of those calculations are being done, but I don’t think they’re being done by the people we’re most likely to talk to, which are usually editors. Like the editors are not the ones doing linear aggression.

[01:06:29.170] – Scott
Oh yeah. And I’m not frustrated with the editors on this point. I’m frustrated with whoever, as you’re saying, whoever is calling the shots there because they do have the data. I’m not convinced they have run these analyses at all because if they had, they would theoretically arrive at this same conclusion that, hey, there’s a minimum viable threshold. That if we don’t get all of our books to that minimum viable threshold, there’s no chance that they could be breakouts and be a surprise success. Right? And in that case, the policy and maybe this is where people are moving, publishers are moving. The corporate policy that makes sense in that case is just don’t acquire books unless you’re going to put a lot of money into them. It just doesn’t make sense. Well, it’s really not surprising because in my last corporate job, it wasn’t my job to be doing things like statistical analysis. I just happened to be trained in it. And I remember one time this is maybe I’ll have to obfuscate some of this a little bit so I don’t get in actual legal trouble. I came back from vacation one time and there was this whole huge issue with a major metric that the company I worked for followed religiously.

[01:07:46.380] – Scott
It was their primary metric on how happy people were with our service. And it was correlated internally to all sorts of different things about how likely big partners were to churn quit our company, basically. And it had gone down by a lot of points all of a sudden. Nobody knew, knew why. Everybody’s good at following really basic charts of very basic data, but when shit changes, they are completely lost. And this is like a five $6 billion company with a lot of theoretically smart people that were just lost. And all I had to do was linear regression. And you have to dig into the data. You have to make sure your data is good. You have to make sure that you understand where all the data is coming from and what could be a potentially confounding factor and all that kind of stuff. But at the end of the day, there just really aren’t people that take the time to go look at, hey, what kind of activities are we doing? And when we do it, what happens? When we don’t do it, what happens? And mathematically, what does that look like? So it’s not surprising to me that this is like a revelation, but I hope that it becomes less of a revelation because whether it’s because they’re listening to this or they just realize that math can help them, they catch on to how easy this is to actually get some actionable insight.

[01:09:16.170] – Sunyi
So just in case any of the various publishing people who are listening to this actually want to get in contact with you or find you, where would they find you? Online. And generally feel free to plug yourself just so that people know who you are. And yeah, in case they do actually ever want to find you to get some information.

[01:09:33.510] – Kerry
So you can find me on social media in a lot of places at swilua. That’s SW I-L-U-A. It stands for she who is like unto aphrodite. And it’s a joke that went awry from a long, long time ago. I used to teach at BYU, which is a very religious school, and they were always trying to call me sister instead of like, doctor or professor or anything. And that was kind of a problem because in this mormon church, women don’t really have a lot of power, and so sometimes people would be like, you can’t give me a grade, or whatever. And that was always really stressful. So I would always try to say, you have to or you should call me, like, doctor. It was spencer back then, but eventually they always struggled with it. So I started putting it in the syllabus. And then one year and it was a joke. It was definitely meant as a joke. I wrote like, a list in the syllabus, and I was like, you can call me dr. Spencer. You can call me professor spencer. You can call me she, who was likened to aphrodite. And the problem with BYU students is they’re very sincere and they didn’t realize it was a joke, and so they legitimately called me by the full title of she was likened to aphrodite for the whole semester.

[01:10:51.460] – Kerry
And it got so embarrassing at a certain point that I shortened it to swalua, and then it just became like this massive inside joke. And here we are. Gosh, it’s maybe 20 years later. Who knows? So that’s my handle on most social media sites twitter, spoutable, post, wherever, whichever one survives the demise of twitter. There’s a little underscore on instagram and threads at the end of spalua, but I think if you search spillu, you can find it. You can also buy my book, which is called I spoke to you with silence. It’s a collection of nonfiction essays from queer mormons, and I think that’s all. If you want to read or buy a book, and you’re an editor and you want to buy a book that’s about like a plague, then you should talk to my agent and I can give you his email address, just DM me, but no, that’s brilliant.

[01:11:45.370] – Sunyi
Thank you. I was looking forward to a lot. It was really, really interesting. Thank you so much. And now I really want to use your algorithm sometime.

[01:11:53.090] – Scott
I’ve been looking forward to this one for a very long time, and I had no idea you were ex mormon at that point today. Really? So now I really feel like I’ve found a kindred spirit.

[01:12:03.420] – Kerry
Thank you for having me. On. Delightful.

[01:12:06.330] – Sunyi
You’ve been listening to the publishing Radio podcast with Sony, Dean and Scott Drakeford. Tune in next time for more in depth discussion on everything publishing industry. See you later.