What Do Comics Sales Numbers Mean?

It has become downright common to see people — actual, human beings — quoting sales figures in online discussions. Go on Facebook, Twitter, Reddit, anywhere and say how you feel about a book. Any book. Eventually, someone will come by to tell you its ComiChron ranking. They’ll point to Comixology’s download figures, or they’ll reference a YouTuber waving a number over their head.

But what does any of this mean? And should anyone actually care?

Hi, my name is John Galati, and I have an MBA. By which I mean I have a MODOK of Business Analysis. And I’m here to hopefully fix everything that’s wrong with the internet’s board of directors.

The numbers, where they come from, and what they measure

There are four major sources of data on comic book sales.

(ComiChron)

This company measures how many books Diamond Comic Distributors sells direct to retail stores in North America. I’m not legally allowed to say that Diamond has a true monopoly on comics distribution in North America, but I can say that they own every hotel and they won’t let anyone else play the bank or be the race car. Anyway, ComiChron’s numbers (which is to say, the race car’s numbers) represents virtually all of the American comic book wholesale business, gives us a great idea of what comic shops are stocking. However, ComiChron does not measure how many books those comic shops actually sell to customers. Basically, ComiChron is a labor intensive shipping manifest.

Reprint Numbers

This is just like it says on the tin. Every time a book sells out of its run (or near enough), publishers will put it back on the presses. In other words, if a book that shipped 20k units last month goes to reprint almost immediately, we can guess that its sales numbers are close to 20k for the month (more on this soon). We can also bet that the nth printing will likely have a new cover and probably have lower value on the speculator market.

(Bleeding Cool)

Each month Rich Johnston takes an extremely informal poll of various comic book shops across North America, gathering up their top sellers for the month and averaging them into a list. But there’s another problem: Rich is not a scientist and how dare we ask him to be. Bleeding Cool’s sample size is very low, inconsistent, and poorly randomized. Compounding matters, BC’s numbers only provide a ranking, not raw data. This means that Bleeding Cool data is interesting (if you’re into this sorta thing), but isn’t necessarily representative of any wider trend.

(ComiXology)

“Best Selling Digital Comics”

Our only digital figures come each week from ComiXology, starting with its à la carte sales figures. The data is fairly global as ComiXology users are from a wide range of countries. That’s a big plus. And it’s backed by Amazon (a fairly neutral party in this situation. However, in this scenario I think they’re the dog, own all of the utilities and railways, and they’ve stolen all of the “chance cards”).  That said, these sales figures are regularly out of alignment with ComiChron and Bleeding Cool, in part because Comixology A. doesn’t always have comics day-and-date with print and B. it trades on steep discounts, with more than half of the Top 10 rankings being dominated by 50% discounts or more.

“Unlimited”

ComiXology also has a paid subscription service, offering free, unlimited reads of select comics from DC, Marvel, IDW, BOOM!, and more. Unlimited titles are “free,” they get their own rating system, independent from the “Best Selling Digital Comics” above. I cannot begin to guess at relative figures between the two lists over time, let alone the way money changes hands in this system.

Recap: What we know

• High degree of certainty on what comic shops are stocking.
• High degree of certainty on what Comixology is selling.
• Moderate certainty on reprint numbers indicating sales.
• Low certainty on what Bleeding Cool’s retail numbers represent.

What we can infer

We can make three big inferences from the numbers available.

1. We can assume in some places that ComiChron numbers represent how many issues a comic book store hopes to sell. Otherwise known as “expected demand.” This tells us interesting things about what comic store owners will buy, the health of that system, and things both good and bad.
2. We know that in other cases, books have artificially high shipping numbers. This happens when publishers take a book that’s expected to sell well (eg. a foil variant cover) and restrict its availability to people who bought x-number of low-selling books. This figure tells us that the speculator market can go straight to hell.
3. We can watch for reprint numbers as an indicator of positive consumer sales and validating ComiChron’s expected demand figures. It is not, however, a sign that the current print run is completely sold out. The exact voodoo behind reprints is both needlessly involved for our purposes and, in many cases, trade secret to the publisher in question. All we need to know for now is that if a print run of 20,000 books goes to reprint, it’s not a sure indicator it’s sold all 20,000 books yet.

What we don’t know

Let’s pretend you wanted to know exactly how many issues of Immortal Hulk #1 were sold. Here are the impediments to fulfilling that lifelong dream.

Problems with existing data

• ComiChron, Bleeding Cool, and ComiXology don’t track actual customer purchases in raw data. Clearly, we need those.
• Most of our current data is from North America, but as of 2013 over half of all ComiXology sales are outside of the US. And multiple artist interviews suggest it’s the same for print.
• Sources like Bleeding Cool would have to be checked for sample size and consistency vs. randomization.
• ComiChron data would have to be purified against micro and macro forces like shipping delays and damages, returns, borrowing programs, and promotional techniques that artificially inflate shipping data (eg. variant covers locked behind stocking quotas)
• We’d need figures on multiple purchases by a single entity (as in person, family, or company.)
• All data would have to be verified by an outside audit.
• I would need a lot of money to untangle it all.

A partial list of missing data sources

Amazon, Marvel Unlimited, DC Universe, Comixology Unlimited, Kindle, Humble, Panel Syndicate, InStockTrades, Crowdfunding sources like Patreon and Kickstarter, Educational sources like Follet and Scholastic, Barnes & Noble, the entire Library system, and convention sales. Also, hospitals and the military. Basically, everything.

So is our data useless?

This may look bad — and from a scientific standpoint, it is — but the data we have isn’t “nothing.” Quite the opposite, the picture it paints is still interesting. After all, I can’t think of a superhero comic sold in the last 10 years that managed to become a “best seller” without also charting on ComiChron, Bleeding Cool, ComiXology, and reprints.

So there is still value to their information and to referencing it, provided all parties are aware of the blind spots.

But then again, I’m not sure that’s why this data gets used online…

Other things you didn’t know where things

Numbers, by themselves, are useless. I don’t mean to get all “money is just paper” on you here, but it’s honestly important to consider the value of data and it’s relative purpose before we get hung up on contrasting figures. To wit, here’s a pair of concepts that will make you want to punch Adam Smith right in his powdered wig.

The Batman Problem

Q: If one book ships 400,000 books a month and another 50,000, which is the better performing book?

A: Depends who’s on the title.

Every publisher has a lynchpin, a title that sells so well it props up the rest of the line. DC has the Batman line which comfortably ships over 700k units a month like clockwork, Marvel has Spider-Man and Avengers (both routinely over 400k units/mo), Image has Saga, and so on.

These books are so big that the screw up our whole conception of comic sales. For Instance, if the best selling Bat-book were to routinely dip beneath 100k/mo, it would be a cataclysm. For fans, for DC, for the entire industry. As a microcosm of this, when Saga skipped a shipping window it created an industry-wide downturn in sales.

On the other hand, Deathstroke hovers around 20k/mo, a number that represents an enormous success for that character’s publication history.

If Batman were to sustain 400k/mo on a single title, he’d get his creative team removed. (As was the case with Tom King).

If Deathstroke hit 40k, his next solicitation would be for a Hollywood film.

Wormholes

That’s right, the shape of the modern comic book economy is a wormhole.

After the ludicrous success of the MCU, it is the goal of all comic book companies to make more films and television. More pointedly, it is the fervent hope of quite a lot of publishers and creators alike to make the next Avengers: Endgame.

As of my writing this, Endgame has a worldwide box office of $2,750,760,348 and a runtime of 181 minutes. Both of these things are soon to increase, but let’s stick with that right now.

181 minutes is a lot of screen time. To fill it, the movie needed pieces from these books (and many more):

• Infinity Gauntlet
• Infinity
• Peter David’s Hulk
• Matt Fraction’s Iron Man
• Civil War
• Civil War II
• House of M
• Asgardians of the Galaxy
• Nick Spencer’s Captain America
• Secret Empire

That’s more than 30 years of publishing history, I don’t know how many pages, and it’s only a start.

All of those books gather together under their own gravity and transform into something new. A singularity of box office success.

Avengers: Endgame made more than $15 million for every minute of runtime. Or $253,292/second (using today’s numbers).

From there, the film is transformed into merchandise. More than the books could ever make on their own. Enough to fill every Target and every Walmart in every dimension.

The most important metric in comics isn’t “how many issues did it sell” but rather “how many pages can be used in a film?” For instance, the first time Captain America lifted Mjolnir was in Mighty Thor #390 (1988). By today’s standards, it’s not a famous issue and the scene doesn’t last that many pages. And yet it made Endgame more than $45 million. And I’m sure the passion it generated contributed to at least a few tshirt sales.

There are also some major influences on that list (Civil War II, Sam Cap, etc) which were not fan-favorites when published… yet they make bank on celluloid. What do Diamond numbers mean in the face of that?

The question isn’t ComiXology or ComiChron numbers; the question is “how much is Captain Marvel’s haircut from The Rise of Alpha Flight worth?”

Are you saying comic books don’t matter?!

No.

But you said, “sales figures aren’t…”

No, I didn’t. I said we’re missing too much data to draw meaningful conclusions. Also, that outside parties (read: “fans and blogs,” of which I’m both) generally have terrible insight into the marketing and business plans of multi-billion dollar brands. I also said some slanderous things about a dead economist. I know that every one of those things is a shock, but they’re all true. All of them.

Okay, but what I heard was that comics are just a slow beta test for movies.

That’s actually pretty catchy, but not entirely true. At present, comic books are still very much their own thing, with their own rules, and they really need to sell as many books as possible. As fast as possible. To as many people as possible.

As long as Marvel, DC, Image, etc want to be comic book publishers, they need to publish comic books for money. Which means they need to rely on sales figures.

But we sure don’t.

What you can count on

We live in a metrics age. We have all, in our own ways, retreated into the safe brutalism of the quantifiable. I say this not merely as a fan of comics, but as someone who has spent the better part of a decade working for data firms, finance companies, and analytics-obsessed marketing shops. In other words, boring math people.

So I’m here to tell you this: high numbers don’t make a book good, and low numbers don’t make it bad. Of course, by measuring online discourse, we might conclude that numbers are all that matters.

Go to any fansite you want and read an argument. Don’t worry, there will be one. Totilo’s law states that within five contentious exchanges, someone will quote a metric that completely misses the point. Usually a sales figure, a stock price, or profit margin; sometimes a dictionary definition (which is the absolute worst).

The problem with doing this is that, outside of a purely business context, the figure has no meaning. Here’s a mental exercise to prove it.

Name 10 musicians that you would consider “top selling.” Now tell me, how many of them do you hate to your very core? You can do the same with fast food, pro sports, beer, anything. You can also just remove the qualifier “top selling” and increase the seed to 30, you’ll still get there.

High sales don’t mean quality. Low sales don’t mean a lack of it. And increasing sales, popularity, or availability doesn’t confer much of anything to the page.

Quantifying our fandom doesn’t make us any more right, or others anymore wrong. And even if it somehow did, it wouldn’t get us one better book or one thin dime.

So go. Go out and be free. I release you from the tyranny of numbers with that promise that none of this will be on the test.