On the book side, one complication to add to the mix: selling more books isn't always the main goal. Often it's getting the ideas out, with selling books being more than a side-benefit, but not the sole focus. Certainly, for a lot of people, selling more books has pretty marginal benefits given that they're unlikely to clear their advance.
Another angle: once you're on a podcast, is it better to give away as much of the book as you can, even if it makes people a little less likely to buy the book, or is it better to play a little coy and give people an intriguing sense of it but making clear that there's still a lot of juice left to squeeze.
Which I guess comes back to your point -- there are some books that are really good and hearing them discussed on a podcast only scratches the surface of what sitting with the book for several hours will provide. Others, less so...
Really enjoyed this, but I think there's a Moneyball angle worth considering.
Moneyball was never about making a given player better. It was about portfolio construction under a budget constraint. Beane couldn't outbid the Yankees for stars, so he built a team from systematically undervalued players.
The publishing version isn't "how do I sell this book?" It's "what kind of books, by what kind of authors, with what kind of audiences, systematically outperform their advances?" That's a portfolio question, and it has answers.
The opportunity isn't for the big publishers. They're the Yankees. They can buy proven authors and absorb the failures. The opportunity is for small presses, self-publishers, and Substack writers building toward a book. The people with a budget constraint. For them, finding undervalued opportunities is the whole game.
This portfolio approach is a good analogy for publishers because each book is really like a startup (new product to new market in highly uncertain conditions), and in aggregate only 10% of an investor’s portfolio pays off.
It's also good for individuals for non-fiction. You are starting with a portfolio of one book.
-Find an audience that you can reach.
-Find a problem that you can solve for that audience.
-Write the book solves that problem.
-Start reaching that audience with your solution.
-If it really solves the problem, try reaching more of that audience with your now proven solution.
In the startup world, Marty Cagan says your product/book needs to be Valuable, Usable, Feasible, Viable. Meaning that you can write a valuble and applicable book for your target audience and reach them and make money.
I think the very open secret is that, to the extent that book publishing can be money-balled, it’s not going to be with serious nonfiction like you and I trade in. It’s going to be with celebrity, romance, and popular self-help. And as much as I think we should all just aim to write good quality stuff as a matter of integrity and self-dignity, it’s just too easy to point to a plethora of bestselling middling stuff also. But no one would say that we should look to that and aim to be middling but successful (last years’ runaway hit was Let Them Theory). Back to the open secret: if you’re in serious nonfiction, you’re likely to be depending on the hit makers to subsidize your book deal, and the press is giving you a shot but basically expects you not to earn out. There’s no way to moneyball serious nonfiction unless you’re just a charismatic celebrity intellectual already. The places to be a guest that sells books don’t really book nobodies. You have to already be known as someone worth talking to.
There is a simpler explanation (that aligns with your final conclusion). Baseball (and poker) are finite games. There are turns and clearly defined outcomes with highly structured rules. You can compare normalized data because they are abnormally consistent systems.
Real life is not like that.
Real life is infinite games, where a good play changes the rules of the game, where variables cannot be isolated, and therefore it is difficult, if not impossible, to do the simple regressions of Moneyball.
There are strong signals in media, ones that have yielded to testing, but these are mostly grotesque: nudity, violence, and wealth sell. Shock value sells until it doesn’t. And cruelty sells - that is why every narcissist podcaster ends up as a nutbag conservative.
The category error is that because a trivial game, like baseball, can be “solved” with statistics, the world can be. “Real” complex social and organizational systems resist. The most successful investors may use deep modeling, but mostly they use insider information.
Baseball (and poker) are also kind learning environments. Book publishing (and most of the adult world) exists in a wicked learning environment. It’s like the difference between a puzzle and a mystery. A puzzle has a solution and defined number of pieces. A mystery may not have a solution and you don’t know how many pieces might be involved.
My day job is as an analytics consultant for book publishers and from a marketing perspective, the goal is to create a snowball effect. Many promotions in many channels in a concentrated period of time. In the data, we are looking for a lift in sales that align with an uptick in marketing and publicity, but it's difficult to attribute those sales to a single podcast episode when there's also retail promotions, other media, publisher promotions, ads, special sales running simultaneously. Even attribution that should be clear is cloudy. Do links in the show notes go to the publisher's website and is the referrer included or does that traffic show as Direct (no attribution). If the listener enjoyed the podcast then searched for the author's name or book title and came via Search then it's unclear if the source of that demand was listening to the podcast. If the listener chose the show notes' affiliate link to Amazon then publishers see sales, not the source of the sales, so is the host sharing that data with the publisher? Usually no. Does this particular podcast invite several authors from the same publisher, where the data can then be used to see if there is always a lift in sales or demand? Sometimes the invites are too far between to see any relationship. And, what about the listener who then purchases at their local bookstore. Nobody is tracking why they chose that book. And there's nothing to say that the listener didn't also see a social post, an ad, or heard a recommendation from a friend. In general, yes there is sales data but no the attribution is not clear so that leaves publisher looking for directional data: when we do this combination of marketing then we see a lift.
My day job is the AoM Podcast. Been doing it for nearly 20 years. And I strictly interview authors. We include Amazon affiliate links to the book we're interviewing about, so I get a snapshot of how effective a podcast is at moving books. A lot of people might not buy the book through the affiliate link so I can't capture that data.
It's really all over the place how many books sell from an episode. Some episodes we'll see hundreds of book sales via our Amazon affiliate link; other episodes just a few dozen.
Books that tend to sell hundreds of copies are ones that solve a problem for people: how to get healthier, how to improve relationships, etc.
Episodes about history and philosophy or things like that will only sell a few dozen.
Basically, it follows books sales in general.
I think authors can be a bit strategic about which shows they appear on. If you've got a general self-help book, then go on a podcast that has a broad, general audience. If your book is about history, philosophy, economics, psychology (like academic type), stick to shows in those niches. They don't have a large following, but the audience is more likely to buy your book compared an audience that has a general audience.
Some academic type books can do well on a more mass audience if the author is engaging and charismatic and can give regular people a good hook. The host can play a big role in this by asking the right questions, but a lot of podcast hosts aren't very good. I'd check out a podcast before accepting an invite to see how the host is
Academic types tend to be better writers than engaging podcast guests. They've got a voice and face for writing. That's fine. I think it's okay to play to your strengths and you don't have to go on podcasts because everyone says you need to go on podcasts.
Someone in the comments asked what strategy an author should take when the talk about their book on a podcast: 1) basically spill the beans on your book or 2) play coy and leave people wanting more.
We've had a lot of guests try to take tact 2. They also do a lot of "as I say in my book" "you can find this in my book" before every answer they give. Yes, we know this is in your book. The topic of this episode is your book. These aren't very good interviews. They sound stilted and leave the audience frustrated.
Surprisingly, the guests that take tact 1 (spill the beans) often sell a lot of books. I've had guests who basically just summarized everything in their book in an episode and they still sold hundreds of books from that episode. Maybe people would have still bought the book if they took tact 2. The book would have sold well regardless. But maybe the audience thought "Wow, this guy was really useful! I liked what he had to say in this convo. I'm going to buy the book to see if there's more or to refresh myself." Who knows. Tact 1 does provide a better show, though.
Book promotion is a lot like advertising. We know half of it works, just don't know which half.
Maybe the best tact is to be Jesus' sower and just scatter your book far and wide and hope a third of it lands on good soil.
I have to think about this stuff too trying to promote my podcast. What's the best way to do it? Am I wasting my time doing this promotional tactic? It's a wicked problem.
The worst thing to do as an author is to be seen trying to get people to buy (or, note the existence of) your book. Also, you have to do this. A conundrum!
Oh, you can moneyball (i.e. ruin) anything. If moneyball hasn't come to publishing that's probably because, unlike with baseball, it's not clear what to optimize for. You suggest optimizing for "book sales," but which book? It could be the one you're promoting now or the several you'll write later.
Ok. Two things, and I’m speaking here as a writer who has done the promotional thing eight-odd times. 1. Author A going on Podcast X ≠ Author B going on Podcast X. Exposure is not a constant. Performance varies. Performance also varies for each author over time. You click with the host or you don’t. You were hungover. A big and unrelated news story happened to break that day. 2. It’s not lab conditions: you can’t do X TV show and then wait to see if there’s an uptick in orders. Because when you’re doing promo, a broadsheet review may have dropped the same day as your TV bit. Someone you did a Q&A with on Substack puts it live a few hours later. There is no baseline to measure from. So it’s messy, unlike baseball, which is clean. Which thing nudged the dial, the festival appearance, the Booktok mention, the chart position on a popular subreddit? No way to know.
As a book author, Red Sox fan, and one who is pro-Sabermetrics until it ruins everything, I really enjoyed this post! My own discussions with book marketing & publicity folks led to the same confidence-without-data replies you got. FWIW, here's my self-report data point on my own modestly selling book (And because authors must self-promote, here's a link to it: https://www.maryellenmacdonald.com/books/more-than-words And, also may I say, it's well liked by many who read it!!). I think I can see a boost from appearances on popular podcasts, similar to the boost from publication of an op-ed. More niche podcasts and publications haven't seemed to be reflected in sales, but of course I don't know what sales would have been like if I hadn't done those podcasts or other publications.
To answer some of @Brett McKay's points in his reply, the degree to which authors can be coy or spill the beans on a podcast depends a lot on the host. Some hosts are laser focused and informed about the book, allowing authors to choose their spot on the coy-to-beans continuum; some hosts want to talk about other writing I've done, not the book; others are nominally aiming to talk about the book but somehow not getting there amidst other topics they want to talk about. I try to be a good engaging guest throughout, but who knows how this sort of variation affects sales or the degree to which I'm getting my own views out there.
Oh, of course I remember, Paul! I also remember your taking me out to lunch the year before when I was interviewing for the MIT job. You had all sorts of interesting questions just as I was realizing that the tuna or chicken salad sandwich I'd ordered had almonds in it--I'm allergic to almonds! I was distracted from whatever you were saying by trying to figure out whether this was like the various times I'd had a few almonds and been OK, or like that time I first discovered the allergy, when I ended up in urgent care. Fortunately it was the former for our lunch.
Great post, Paul! I wonder what the answer would be... probably like most things in social science: many small effects in exactly the direction you'd predict. For example, all else being equal, probably going on more podcasts is correlated with more book sales (whatever the direction of causality). Or being on a bigger publisher is correlated, or getting earned media etc.
I would bet that the conventional wisdom is right in the sense that it identifies real positive relationships but that those effects are all too small to meaningfully tell you about the fate of any one book or author.
I think you're right that the mechanics of book publishing/buying are too variable to be Moneyballed. Ultimately, if a book is going to truly break out, it must become a "word-of-mouth" hit as readers tell others that this is something worth their time. Getting on Oprah or Fresh Air or whatever can increase the odds that something will achieve word-of-mouth status, but it is no guarantee. It's interesting to see how varied sales can be for books selected by Reese and Jenna (the two most powerful book clubs). There is a high floor relative to the overwhelming majority of books published, but the ceiling really varies. The ones that reach the top of the distribution achieve word-of-mouth.
Given the incredible promotion Isgur's book received, the available information says that she's not yet achieving word of mouth success. A spike on the bestseller list after that kind of burst of publicity and having a core audience through your own podcast is not unexpected, but it has not persisted. This isn't unusual for nonfiction. In any given week the Times nonfiction list is mostly first-time entries and the books that persist for more than one or two weeks are rare.
Anecdotally, I can report that in a handful of cases listening to an author on a podcast talk about their book has satisfied me to the point where I did not feel the need to buy the book. The most recent was Belle Burden's "Strangers" where the two interviews I experienced seemed to tell me all I needed to know. Powerful story, but I didn't have to read it.
I'm a book person. I can't afford to buy every book that sounds interesting. Even if I could afford them, I don't have the bookshelves for them - or the wall space for more bookshelves. A first option is finding the book at a library, but many of the books that interest me are not the sort local libraries acquire. My next step is to look for a podcast interview with the author. Sometimes the interview tells me all I need to know from the book. After all, I can't afford all of them. Other times the interview confirms that yes, this is a book I should acquire for myself.
On the book side, one complication to add to the mix: selling more books isn't always the main goal. Often it's getting the ideas out, with selling books being more than a side-benefit, but not the sole focus. Certainly, for a lot of people, selling more books has pretty marginal benefits given that they're unlikely to clear their advance.
Another angle: once you're on a podcast, is it better to give away as much of the book as you can, even if it makes people a little less likely to buy the book, or is it better to play a little coy and give people an intriguing sense of it but making clear that there's still a lot of juice left to squeeze.
Which I guess comes back to your point -- there are some books that are really good and hearing them discussed on a podcast only scratches the surface of what sitting with the book for several hours will provide. Others, less so...
Really enjoyed this, but I think there's a Moneyball angle worth considering.
Moneyball was never about making a given player better. It was about portfolio construction under a budget constraint. Beane couldn't outbid the Yankees for stars, so he built a team from systematically undervalued players.
The publishing version isn't "how do I sell this book?" It's "what kind of books, by what kind of authors, with what kind of audiences, systematically outperform their advances?" That's a portfolio question, and it has answers.
The opportunity isn't for the big publishers. They're the Yankees. They can buy proven authors and absorb the failures. The opportunity is for small presses, self-publishers, and Substack writers building toward a book. The people with a budget constraint. For them, finding undervalued opportunities is the whole game.
Good point, Mark -- yes, I see that there are pretty different things.
This portfolio approach is a good analogy for publishers because each book is really like a startup (new product to new market in highly uncertain conditions), and in aggregate only 10% of an investor’s portfolio pays off.
It's also good for individuals for non-fiction. You are starting with a portfolio of one book.
-Find an audience that you can reach.
-Find a problem that you can solve for that audience.
-Write the book solves that problem.
-Start reaching that audience with your solution.
-If it really solves the problem, try reaching more of that audience with your now proven solution.
In the startup world, Marty Cagan says your product/book needs to be Valuable, Usable, Feasible, Viable. Meaning that you can write a valuble and applicable book for your target audience and reach them and make money.
That's Moneyball for book writers.
I think the very open secret is that, to the extent that book publishing can be money-balled, it’s not going to be with serious nonfiction like you and I trade in. It’s going to be with celebrity, romance, and popular self-help. And as much as I think we should all just aim to write good quality stuff as a matter of integrity and self-dignity, it’s just too easy to point to a plethora of bestselling middling stuff also. But no one would say that we should look to that and aim to be middling but successful (last years’ runaway hit was Let Them Theory). Back to the open secret: if you’re in serious nonfiction, you’re likely to be depending on the hit makers to subsidize your book deal, and the press is giving you a shot but basically expects you not to earn out. There’s no way to moneyball serious nonfiction unless you’re just a charismatic celebrity intellectual already. The places to be a guest that sells books don’t really book nobodies. You have to already be known as someone worth talking to.
You reminded me of Lord Leverhulme’s comment that "Half my advertising spend is wasted; the trouble is, I don't know which half."
There is a simpler explanation (that aligns with your final conclusion). Baseball (and poker) are finite games. There are turns and clearly defined outcomes with highly structured rules. You can compare normalized data because they are abnormally consistent systems.
Real life is not like that.
Real life is infinite games, where a good play changes the rules of the game, where variables cannot be isolated, and therefore it is difficult, if not impossible, to do the simple regressions of Moneyball.
There are strong signals in media, ones that have yielded to testing, but these are mostly grotesque: nudity, violence, and wealth sell. Shock value sells until it doesn’t. And cruelty sells - that is why every narcissist podcaster ends up as a nutbag conservative.
The category error is that because a trivial game, like baseball, can be “solved” with statistics, the world can be. “Real” complex social and organizational systems resist. The most successful investors may use deep modeling, but mostly they use insider information.
Baseball (and poker) are also kind learning environments. Book publishing (and most of the adult world) exists in a wicked learning environment. It’s like the difference between a puzzle and a mystery. A puzzle has a solution and defined number of pieces. A mystery may not have a solution and you don’t know how many pieces might be involved.
I think a puzzle is a finite game, and real people, especially those who excel in business, are often infinitely diverse in their frustratingness.
My day job is as an analytics consultant for book publishers and from a marketing perspective, the goal is to create a snowball effect. Many promotions in many channels in a concentrated period of time. In the data, we are looking for a lift in sales that align with an uptick in marketing and publicity, but it's difficult to attribute those sales to a single podcast episode when there's also retail promotions, other media, publisher promotions, ads, special sales running simultaneously. Even attribution that should be clear is cloudy. Do links in the show notes go to the publisher's website and is the referrer included or does that traffic show as Direct (no attribution). If the listener enjoyed the podcast then searched for the author's name or book title and came via Search then it's unclear if the source of that demand was listening to the podcast. If the listener chose the show notes' affiliate link to Amazon then publishers see sales, not the source of the sales, so is the host sharing that data with the publisher? Usually no. Does this particular podcast invite several authors from the same publisher, where the data can then be used to see if there is always a lift in sales or demand? Sometimes the invites are too far between to see any relationship. And, what about the listener who then purchases at their local bookstore. Nobody is tracking why they chose that book. And there's nothing to say that the listener didn't also see a social post, an ad, or heard a recommendation from a friend. In general, yes there is sales data but no the attribution is not clear so that leaves publisher looking for directional data: when we do this combination of marketing then we see a lift.
My day job is the AoM Podcast. Been doing it for nearly 20 years. And I strictly interview authors. We include Amazon affiliate links to the book we're interviewing about, so I get a snapshot of how effective a podcast is at moving books. A lot of people might not buy the book through the affiliate link so I can't capture that data.
It's really all over the place how many books sell from an episode. Some episodes we'll see hundreds of book sales via our Amazon affiliate link; other episodes just a few dozen.
Books that tend to sell hundreds of copies are ones that solve a problem for people: how to get healthier, how to improve relationships, etc.
Episodes about history and philosophy or things like that will only sell a few dozen.
Basically, it follows books sales in general.
I think authors can be a bit strategic about which shows they appear on. If you've got a general self-help book, then go on a podcast that has a broad, general audience. If your book is about history, philosophy, economics, psychology (like academic type), stick to shows in those niches. They don't have a large following, but the audience is more likely to buy your book compared an audience that has a general audience.
Some academic type books can do well on a more mass audience if the author is engaging and charismatic and can give regular people a good hook. The host can play a big role in this by asking the right questions, but a lot of podcast hosts aren't very good. I'd check out a podcast before accepting an invite to see how the host is
Academic types tend to be better writers than engaging podcast guests. They've got a voice and face for writing. That's fine. I think it's okay to play to your strengths and you don't have to go on podcasts because everyone says you need to go on podcasts.
Someone in the comments asked what strategy an author should take when the talk about their book on a podcast: 1) basically spill the beans on your book or 2) play coy and leave people wanting more.
We've had a lot of guests try to take tact 2. They also do a lot of "as I say in my book" "you can find this in my book" before every answer they give. Yes, we know this is in your book. The topic of this episode is your book. These aren't very good interviews. They sound stilted and leave the audience frustrated.
Surprisingly, the guests that take tact 1 (spill the beans) often sell a lot of books. I've had guests who basically just summarized everything in their book in an episode and they still sold hundreds of books from that episode. Maybe people would have still bought the book if they took tact 2. The book would have sold well regardless. But maybe the audience thought "Wow, this guy was really useful! I liked what he had to say in this convo. I'm going to buy the book to see if there's more or to refresh myself." Who knows. Tact 1 does provide a better show, though.
Book promotion is a lot like advertising. We know half of it works, just don't know which half.
Maybe the best tact is to be Jesus' sower and just scatter your book far and wide and hope a third of it lands on good soil.
I have to think about this stuff too trying to promote my podcast. What's the best way to do it? Am I wasting my time doing this promotional tactic? It's a wicked problem.
Thanks, Brett -- that's very interesting. (I like the Jesus analogy)
The worst thing to do as an author is to be seen trying to get people to buy (or, note the existence of) your book. Also, you have to do this. A conundrum!
Oh, you can moneyball (i.e. ruin) anything. If moneyball hasn't come to publishing that's probably because, unlike with baseball, it's not clear what to optimize for. You suggest optimizing for "book sales," but which book? It could be the one you're promoting now or the several you'll write later.
Ok. Two things, and I’m speaking here as a writer who has done the promotional thing eight-odd times. 1. Author A going on Podcast X ≠ Author B going on Podcast X. Exposure is not a constant. Performance varies. Performance also varies for each author over time. You click with the host or you don’t. You were hungover. A big and unrelated news story happened to break that day. 2. It’s not lab conditions: you can’t do X TV show and then wait to see if there’s an uptick in orders. Because when you’re doing promo, a broadsheet review may have dropped the same day as your TV bit. Someone you did a Q&A with on Substack puts it live a few hours later. There is no baseline to measure from. So it’s messy, unlike baseball, which is clean. Which thing nudged the dial, the festival appearance, the Booktok mention, the chart position on a popular subreddit? No way to know.
As a book author, Red Sox fan, and one who is pro-Sabermetrics until it ruins everything, I really enjoyed this post! My own discussions with book marketing & publicity folks led to the same confidence-without-data replies you got. FWIW, here's my self-report data point on my own modestly selling book (And because authors must self-promote, here's a link to it: https://www.maryellenmacdonald.com/books/more-than-words And, also may I say, it's well liked by many who read it!!). I think I can see a boost from appearances on popular podcasts, similar to the boost from publication of an op-ed. More niche podcasts and publications haven't seemed to be reflected in sales, but of course I don't know what sales would have been like if I hadn't done those podcasts or other publications.
To answer some of @Brett McKay's points in his reply, the degree to which authors can be coy or spill the beans on a podcast depends a lot on the host. Some hosts are laser focused and informed about the book, allowing authors to choose their spot on the coy-to-beans continuum; some hosts want to talk about other writing I've done, not the book; others are nominally aiming to talk about the book but somehow not getting there amidst other topics they want to talk about. I try to be a good engaging guest throughout, but who knows how this sort of variation affects sales or the degree to which I'm getting my own views out there.
Thanks, Maryellen. Good points. (I would find it impossible to be coy in a podcast--it seems impolite to both the podcast host and the audience).
(You won't remember this, by the way, but you were a new assistant prof at MIT just as I was in my last year of grad school there.)
Oh, of course I remember, Paul! I also remember your taking me out to lunch the year before when I was interviewing for the MIT job. You had all sorts of interesting questions just as I was realizing that the tuna or chicken salad sandwich I'd ordered had almonds in it--I'm allergic to almonds! I was distracted from whatever you were saying by trying to figure out whether this was like the various times I'd had a few almonds and been OK, or like that time I first discovered the allergy, when I ended up in urgent care. Fortunately it was the former for our lunch.
If you ask people a question they feel obligated to come up with an answer. If they can it will be one that passes the work back to you.
Appearing on Oprah, of course, makes you a celeb so it promotes you not the book.
Most podcasts just promote the book.
Great post, Paul! I wonder what the answer would be... probably like most things in social science: many small effects in exactly the direction you'd predict. For example, all else being equal, probably going on more podcasts is correlated with more book sales (whatever the direction of causality). Or being on a bigger publisher is correlated, or getting earned media etc.
I would bet that the conventional wisdom is right in the sense that it identifies real positive relationships but that those effects are all too small to meaningfully tell you about the fate of any one book or author.
I think you're right that the mechanics of book publishing/buying are too variable to be Moneyballed. Ultimately, if a book is going to truly break out, it must become a "word-of-mouth" hit as readers tell others that this is something worth their time. Getting on Oprah or Fresh Air or whatever can increase the odds that something will achieve word-of-mouth status, but it is no guarantee. It's interesting to see how varied sales can be for books selected by Reese and Jenna (the two most powerful book clubs). There is a high floor relative to the overwhelming majority of books published, but the ceiling really varies. The ones that reach the top of the distribution achieve word-of-mouth.
Given the incredible promotion Isgur's book received, the available information says that she's not yet achieving word of mouth success. A spike on the bestseller list after that kind of burst of publicity and having a core audience through your own podcast is not unexpected, but it has not persisted. This isn't unusual for nonfiction. In any given week the Times nonfiction list is mostly first-time entries and the books that persist for more than one or two weeks are rare.
Anecdotally, I can report that in a handful of cases listening to an author on a podcast talk about their book has satisfied me to the point where I did not feel the need to buy the book. The most recent was Belle Burden's "Strangers" where the two interviews I experienced seemed to tell me all I needed to know. Powerful story, but I didn't have to read it.
I'm a book person. I can't afford to buy every book that sounds interesting. Even if I could afford them, I don't have the bookshelves for them - or the wall space for more bookshelves. A first option is finding the book at a library, but many of the books that interest me are not the sort local libraries acquire. My next step is to look for a podcast interview with the author. Sometimes the interview tells me all I need to know from the book. After all, I can't afford all of them. Other times the interview confirms that yes, this is a book I should acquire for myself.