The Volokh Conspiracy - Tinkerbell Returns!

A good while ago here on the VC (in 2003, to be precise) I introduced (and subsequently named) the "Tinkerbell Effect" (and its close cousin, the "Reverse Tinkerbell Effect"). A Tinkerbell occurs when some phenomenon is more likely to become true simply as a consequence of more people believing it to be true; a Reverse Tinkerbell occurs when some phenomenon is more likely to become false simply as a consequence of more people believing it to be true. [And a shout-out to VC reader Patick Hynes, who helped me, back then, find the name for the phenomenon -- derived, of course, from the moment in Peter Pan when little Tink's light starts to dim, and Peter explains to the audience that Tink is ill, and might just die, because nobody believes in fairies anymore, at which point he exhorts the audience to believe in fairies, and to clap and to cry out "I believe!" -- after which Tinkerbell recovers and everyone feels wonderful. [It always worked, too!] The Reverse Tinkerbell, then, occurs when the more you believe in something, the more likely it is to vanish.]

[Some examples of Reverse Tinkerbells: Voting: The more that people believe the truth of the proposition "My vote matters," the less likely it is that it will become true (because more people will vote if they believe it matters, and that makes it less likely that your vote matters); similarly, the more that people believe the proposition "My vote doesn't matter," the less likely it is to become true (i.e., fewer people will vote, and then your vote will matter). Or: The more that people believe the proposition " X [the college library; Cancun; etc.] is a quiet and unspoiled spot," the less true it becomes. Or: The more people think that some charity (e.g., the March of Dimes, or the Salvation Army) really needs money, the less money it will need.]

Tinkerbells and Reverse Tinkerbells are everywhere, once you start looking for them. Here's one that's been on my mind recently. Anyone who spends 15 minutes or more thinking seriously about how to market a book in this country to a wide audience (as I have been doing of late) realizes that the proposition "Book X is a New York Times Bestseller" is a classic Tinkerbell -- the more people believe the proposition, the more likely it is to become true. That is, of course, why publishers will splash the fact (if it's true) all over their advertisements, and all over the covers of their books (when the new editions come out), and all over their marketing material, etc. [The actual links connecting the belief in the proposition and the outcome are quite interesting and could use some careful study, I think -- once a book is a NY Times Bestseller, many things happen: NPR is more likely to carry a story about it; Terri Gross is more likely to interview the author; bookstores all over the country are more likely to feature it on their display shelves; ordinary consumers are more likely to shell out their hard-earned money (hey - it's a NY Times Bestseller, how bad can it be?...)]

That's all pretty commonplace - like I said, everyone who thinks about selling books understands this. Like every other author out there, I'd love to figure out a way to ride this Tinkerbell to greater and greater heights, to get those positive feedback belief loops humming. How to do that? Well, I could just put "In Search of Jefferson's Moose: a New York Times Bestseller!!" all over the webpage for my book, the Amazon page for my book, the Google AdWords campaign for my book, etc. and watch my sales go up.

Unfortunately, there's a name for that: "fraud."

But here's the interesting wrinkle. On the Net, these loops can aggregate and amplify these signals with astonishing speed [see, e.g., "marketing, viral"], and it might be possible to game this system while avoiding serious exposure for fraud. Here's the idea: assume (even if you don't happen to believe that it's true in this instance) that this particular Tinkerbell really can induce increased sales of Book X. For a fairly modest outlay of money -- I'd guess $20 or $30K could do it -- one could design a blitz Internet ad campaign, saturating Google's AdWords, Amazon and bn.com and maybe some of the other online retail sites, and some of the big blogs, getting that message in front of hundreds of thousands or millions of eyeballs. The ad campaign focuses on the message: "Book X: A New York Times Bestseller!!" The question is: can you get enough velocity to actually make that become true soon enough so that nobody feels "defrauded" or cheated by the transaction? That is, if you were reasonably confident that you could actually generate enough sales on Monday, Tuesday, and Wednesday to make the book an actual NYT bestseller list on the following Sunday's list, who will claim to have been defrauded? Indeed, will there have actually been fraud, in that circumstance?

It probably wouldn't actually work in this instance -- this particular market's not quite friction-less enough to induce it to happen quickly enough; it takes time for realspace booksellers to reorganize their shelves, and time for Terri Gross and Jon Stewart to put together their lineups, etc. etc. My guess, though, is that someone, surely, is going to try. [Not me, by the way -- it sounds too much like Bernard Madoff for my tastes. Rest assured: if you see an ad for my book trumpeting that it's a NY Times Bestseller, the claim is factually true. [And those of you who are more devious than I have already noticed that my getting you to believe that could just be the linchpin of my Tinkerbell strategy . . .]].

And here's a final nice touch -- the belief in the efficacy of this kind of scheme to game the NY Times Bestseller System is itself a Reverse Tinkerbell! The more people believe the truth of the proposition "The NYT Bestseller System can be gamed," the less likely it is to become true (because if people believe that system can be gamed, the "New York Times Bestseller!!" label no longer induces people to buy the book . . .)

	[David Post, March 11, 2009 at 12:23pm] Trackbacks Tinkerbell Returns! A good while ago here on the VC (in 2003, to be precise) I introduced (and subsequently named) the "Tinkerbell Effect" (and its close cousin, the "Reverse Tinkerbell Effect"). A Tinkerbell occurs when some phenomenon is more likely to become true simply as a consequence of more people believing it to be true; a Reverse Tinkerbell occurs when some phenomenon is more likely to become false simply as a consequence of more people believing it to be true. [And a shout-out to VC reader Patick Hynes, who helped me, back then, find the name for the phenomenon -- derived, of course, from the moment in Peter Pan when little Tink's light starts to dim, and Peter explains to the audience that Tink is ill, and might just die, because nobody believes in fairies anymore, at which point he exhorts the audience to believe in fairies, and to clap and to cry out "I believe!" -- after which Tinkerbell recovers and everyone feels wonderful. [It always worked, too!] The Reverse Tinkerbell, then, occurs when the more you believe in something, the more likely it is to vanish.] [Some examples of Reverse Tinkerbells: Voting: The more that people believe the truth of the proposition "My vote matters," the less likely it is that it will become true (because more people will vote if they believe it matters, and that makes it less likely that your vote matters); similarly, the more that people believe the proposition "My vote doesn't matter," the less likely it is to become true (i.e., fewer people will vote, and then your vote will matter). Or: The more that people believe the proposition " X [the college library; Cancun; etc.] is a quiet and unspoiled spot," the less true it becomes. Or: The more people think that some charity (e.g., the March of Dimes, or the Salvation Army) really needs money, the less money it will need.] Tinkerbells and Reverse Tinkerbells are everywhere, once you start looking for them. Here's one that's been on my mind recently. Anyone who spends 15 minutes or more thinking seriously about how to market a book in this country to a wide audience (as I have been doing of late) realizes that the proposition "Book X is a New York Times Bestseller" is a classic Tinkerbell -- the more people believe the proposition, the more likely it is to become true. That is, of course, why publishers will splash the fact (if it's true) all over their advertisements, and all over the covers of their books (when the new editions come out), and all over their marketing material, etc. [The actual links connecting the belief in the proposition and the outcome are quite interesting and could use some careful study, I think -- once a book is a NY Times Bestseller, many things happen: NPR is more likely to carry a story about it; Terri Gross is more likely to interview the author; bookstores all over the country are more likely to feature it on their display shelves; ordinary consumers are more likely to shell out their hard-earned money (hey - it's a NY Times Bestseller, how bad can it be?...)] That's all pretty commonplace - like I said, everyone who thinks about selling books understands this. Like every other author out there, I'd love to figure out a way to ride this Tinkerbell to greater and greater heights, to get those positive feedback belief loops humming. How to do that? Well, I could just put "In Search of Jefferson's Moose: a New York Times Bestseller!!" all over the webpage for my book, the Amazon page for my book, the Google AdWords campaign for my book, etc. and watch my sales go up. Unfortunately, there's a name for that: "fraud." But here's the interesting wrinkle. On the Net, these loops can aggregate and amplify these signals with astonishing speed [see, e.g., "marketing, viral"], and it might be possible to game this system while avoiding serious exposure for fraud. Here's the idea: assume (even if you don't happen to believe that it's true in this instance) that this particular Tinkerbell really can induce increased sales of Book X. For a fairly modest outlay of money -- I'd guess $20 or $30K could do it -- one could design a blitz Internet ad campaign, saturating Google's AdWords, Amazon and bn.com and maybe some of the other online retail sites, and some of the big blogs, getting that message in front of hundreds of thousands or millions of eyeballs. The ad campaign focuses on the message: "Book X: A New York Times Bestseller!!" The question is: can you get enough velocity to actually make that become true soon enough so that nobody feels "defrauded" or cheated by the transaction? That is, if you were reasonably confident that you could actually generate enough sales on Monday, Tuesday, and Wednesday to make the book an actual NYT bestseller list on the following Sunday's list, who will claim to have been defrauded? Indeed, will there have actually been fraud, in that circumstance? It probably wouldn't actually work in this instance -- this particular market's not quite friction-less enough to induce it to happen quickly enough; it takes time for realspace booksellers to reorganize their shelves, and time for Terri Gross and Jon Stewart to put together their lineups, etc. etc. My guess, though, is that someone, surely, is going to try. [Not me, by the way -- it sounds too much like Bernard Madoff for my tastes. Rest assured: if you see an ad for my book trumpeting that it's a NY Times Bestseller, the claim is factually true. [And those of you who are more devious than I have already noticed that my getting you to believe that could just be the linchpin of my Tinkerbell strategy . . .]]. And here's a final nice touch -- the belief in the efficacy of this kind of scheme to game the NY Times Bestseller System is itself a Reverse Tinkerbell! The more people believe the truth of the proposition "The NYT Bestseller System can be gamed," the less likely it is to become true (because if people believe that system can be gamed, the "New York Times Bestseller!!" label no longer induces people to buy the book . . .)