The exact sales of digital comics are somewhat shrouded in mystery.  We’ve been assured that _most_ DC and Marvel titles sell within 10%-15% of the print totals and that the sales rank/order is somewhat consistent, with some notable exceptions.  It’s been offered that independent titles might have higher sales, percentage-wise… at least here and there.

So what would happen if we took Comixology’s top 20 sellers and compared them to the sales charts?  This isn’t a _perfect_ experiment.  Comixology doesn’t have a monthly chart and the most recent sales estimates are from December, not January… but it’s an interesting way to test that conventional wisdom of digital and print sales mostly lining up in the same order.

So let’s see how it looks.  On the left will be the Comixology sales rank as recorded on Sunday evening, 1/21/18.  On the right will the the previous issue (or closest to it if it’s already double shipped in January) from the Comichron.com December sales estimates.

  1. Batman #39 – #38 was at 92,902
  2. Super Sons #12 – #11 was at 37,356
  3. The Might Thor #703 – #702 was at 41,581
  4. Superman #39 – #38 was at 44,281
  5.  Justice League #37 – #35 was at 46,228
  6. Star Wars #42 – #40 was at 61,064
  7. Injustice 2 #41 – N/A Digital First Title
  8. Avengers #676 – #674 was at 33,199
  9. Nightwing #37 – #35 was at 29,310
  10. X-Men Gold #20 – #18 was at  37,616
  11. Trinity #17 – #16 was at 25,178
  12. Doctor Strange #384 – #383 was at 23,021
  13. Aquaman #32 – #31 was at 25,474
  14. Damage #1 – N/A
  15. Green Lanterns #39 – #37 was at 26,055
  16. All-New Wolverine #29 – #28 was at 29,512
  17. Champions #16 – #15 was at 23,190
  18. Kill or Be Killed #15 – #14 was at 16,36 (November)
  19. Amazing Spider-Man Venom, Inc. Omega #1 – Venom, Inc. Alpha #1 was at 69,600
  20. Peter Parker: The Spectacular Spider-Man #299 – #298 was at 29,431

What can we learn from pairing this sample with some print sales data?  Mostly, it opens up more mysteries, although it does suggest that digital sales might not be the predictable percentage they used to be.

The biggest question is whether Star Wars is under performing in digital or Super Sons, The Mighty Thor and Superman are all performing much better than you might expect?

OK, any way that you look at it, Super Sons appears to be performing a bit more strongly in digital, relative to the other titles.  Kill or Be Killed is also showing up in the company of titles that sell better in print.  Champions might be a little higher, too.  Super Sons and Champions might be evidence to support the old theory that some of the more directly YA-targeted material performs more strongly in digital, at least in terms of serialization.

It’s definitely interesting to see Venom, Inc. so much lower on the chart… but then again, you don’t have variant covers propping things up with digital.  Digital is for reading.

Damage popping up between Aquaman and Green Lanterns doesn’t speak to much pent up demand for that title with the digital audience, but perhaps it will fare better with the Direct Market audience.

It’s just one data sample and things may look quite different taken across an entire month, but the sales ranking sure does seem to have more than slight differences from the print numbers we see every month.

Want to learn more about how comics publishing and digital comics work?  Try Todd’s book, Economics of Digital Comics

 

17 COMMENTS

  1. It would be a big help if we knew over what timeframe the Comixology ranking was measuring. Best seller over the last six hours? Over the last week? I tend to think it is at least a few days or it would look more random than it is, but it would sure help to know.

  2. The apples/oranges quotient here isn’t very good.

    You are comparing December hard-copy sales to January online sales… you want to do this analysis by saving the January data from Comixology that you now have and compare it to the January Diamond/ICv2 data that will come out in February. There is no guarantee that the things that drove shops to order December issues of comics are going to have any relationship to the things that motivate customers to purchase January issues from Comixology (which is at least part of what Rich Johnston is alluding to, I think).

    JJM’s comment is also very relevant here.

    Also, when Comixology says “purchases”, do they include free code redemptions? In other words, is the Comixology list based only on *paid* purchases of digital content? I think most people would assume that the answer is “yes” but in fact we don’t know that for sure, do we?

  3. Hi Todd

    Interesting, but I don’t think this is enough data to draw any conclusions. Do you intend to repeat this feature?

    The Android App list is a different order to the Website list e.g. looking right now Superman is 3rd on app, 5th website. I’ve been keeping an eye on bestsellers list on Comixology (though not recording it) and don’t believe it is simply timing. Saga volume 8 was consistently top 10 website, and not continuosly top 25 in the first two weeks after release. Also on the app Reed Gunther #13 has been stuck in top 25 for months so I don’t trust the algorithms. I don’t have access to an Apple device, or really know how kindle plays into it or if they show different orders.

    Do we have a reliable source that gives 10-15% digital? I’ve often seen 10% quoted, but I’ve never seen the source and am unconvinced based on the number of ratings. To take an example you’ve stated Batman #38 as 92,902 sales (so 9,290 to 13,935 digital assumed) but as of writing this has only 224 ratings. This would mean 2.4% to 1.6% of people who bought Batman #38 rated it. The rating screen comes up when you finish an issue and I just feel like far more than 1 in 40 or 60 people (especialy comic fans) would give a star rating.

    @John Jackson Miller – I think it is an algorithm that highly focuses on the last 7 days. It updates at least daily and each Wed/Thur list is dominated by the new releases. Good sales sometimes getsthose discounted comics in the list.

    @ Richard Johnston. I don’t think Diamond lists being to retailers has TOO much impact as retailers buy based on what they think they can sell. Todd’s experiment isn’t apples to apples as it uses figures from previous month issues’s sales, and takes a snapshot at one day to compare to a month.

  4. “There is no guarantee that the things that drove shops to order December issues of comics are going to have any relationship to the things that motivate customers to purchase January issues from Comixology…”

    Just to clarify a little bit:

    Retailer incentives (such as order threshold variant covers) for a title can change from month to month. These would not be expected to have much (if any) influence on Comixology purchases. So incentives in place for a December shipping comic could lead to a dissonance with the Comixology purchase patterns.

    Also, changes in creators and storylines in January as compared to December releases may mean that reader purchasing is different in January vs. December. Similarly, “standard attrition” between December and January also cause a decrease in ordering in January as compared to December. To the extent that standard attrition happens on Comixology, you will have absorbed the January attrition in the digital sales placement but not in the Diamond hard copy data because that is from December.

    So you really need to compare January digital numbers with January Diamond data.

  5. @comicsatemybrain

    I do agree, but then we’d have had to wait an extra month! In reality the data from comixology just isn’t enough to do any analysis for all the reasons stated, so January to January wouldn’t be precise either.

    As John Jackson Miller raised, this isn’t a month’s data, but a snapshot at a moment in time and not clear over what period calculated.

    Also, it really depends on consumer buying habits – how many readers buy week of release – 10%, 50%, 90%? Perhaps they buy every month on payday, or when comics reduce in price? This would spread the digital sales and potentially disconnect the snapshots* from monthly charts. Retailers will buy all the issues they think they can sell in month of release.

    *I assume I’m right that snapshots is mostly or entirely based on last 7 days.

  6. The 10-15% digital figure is from my reports with Milton Griepp over the years, who’s been calculating a report on digital since being a Comixology board member near the very beginning. It does not account for the all-you-can-eat services.

    I continue to think that the number of digital units is an interesting data point to the degree that it expresses audience size, but that a digital unit and a comic book are not interchangeable in their impact on the business. Most obviously, a physical comic book is resalable, adding to the several hundred millions of dollars that the backmarket adds to the business annually. I also tend to believe print success would be the better predictor of collected-edition success (because aftermarket activity multiplies the mindshare or footprint of the IP), but there’s no easy way to quantify that.

    A digital figure I would be greatly interested in: What’s the churn rate in the readership? Are digital readers more likely or less likely to lapse? I tend to think there’s arguments in both directions: obviously the digital reader faces fewer barriers in keeping up a subscription, but the print buyer also may be the more devoted reader for that very reason.

  7. Replying to several questions:
    * I’ve have publishers privately use the 10%-15% figure when discussing single issue digital sales. Some are more likely to say 10% than others. Which has been consistent with Milt’s information. (We may also have spoken with the same folks, so it may or may not be independent confirmation, but I think we can be confident that’s the range to use as a rule of thumb.)
    * If I had to place a wager on the time period of Comixology’s ranking, I’d say that there’s probably either a reset or large change in the weight of sales that coincides with new releases on Wednesday. While you’ll see issues from a previous week near the top from time to time, you normally see a wholesale changeover for the week’s new issues all at once. Still, we really don’t have much information on the extent to which readers pick up new/new-ish issues after the release week. It might be significant and it might not.
    * _Absolutely_ this is a reader list, not a sell-in list. That’s part of why looking at the ranking of variant-intensive issues is so fascinating. And if a lot of Marvel titles fall lower on the digital list, that could well be an indication of how the variants are padding raw sales over actual number of readers.
    *If there’s reader interest, I’ll continue this experiment. Currently, it looks like there’s interest.
    *Yes, comparing the week’s ranking to the eventual January estimates would be less apples to oranges, but this is a first step. The weekly (we think, anyway) nature of the list is always going to introduce some added ambiguity, but if trends emerge, we can likely still learn something from this.

  8. @Ryan —

    I agree with all of your caveats. No doubt about it, they are all important.

    But my point remains that additional unnecessary problems with the data comparison are introduced when comparing data about physical comics released in December with data about the digital purchases of the subsequent issues.

    Yes, I fully agree with you that January to January comparisons still will have problems, making the analysis murky at best. But December to January comparisons have even more problems piled on top of that, but those can be mitigated by switching to January to January analysis. (If the internet existed back in 1983, can you imagine using this analysis to compare shipping estimates of Thor #336 with digital purchases of Thor #337?)

  9. Amy Ziegfeld has been collecting the Top 25s for a couple of years now, and I’ve been consulting them from time to time. The problem is that, even more so than the print reorder lists, it’s a gumbo of ingredients, and mapping it to in-store results is hard.

    For example, take December’s Top 5 at retail. The order of rankings is

    1) DOOMSDAY CLOCK #2
    2) DARK NIGHTS METAL #4
    3) PHOENIX RESURRECTION #1

    which are all clustered within 10% of one another. There’s a big drop to BATMAN 36-37, which are 4-5.

    These landed in Amy’s weekly Comixology checks at:

    DOOMSDAY CLOCK #2:
    12/27: 1st

    DARK NIGHTS METAL #4:
    12/20: 2nd

    PHOENIX RESURRECTION #1:
    12/27: 6th

    BATMAN #36:
    12/6: 2nd

    BATMAN #37:
    12/20: 1st

    …and none of them repeating, as she only lists the top 15 or so per publisher. (No issue numbers either, but it’s something you can figure out usually.)

    Meanwhile, the first-place Comixology books of 12/6 — WALKING DEAD #174 — and 12/13 — MISTER MIRACLE #5 — landed at 12th and 34th respectively at Diamond in December.

    Given the above facts, I don’t know how one would use them to predict that the “top seeds” digitally from the four weeks would wind up 12th, 34th, 4th, and 1st respectively for the month. There’s some correlation, but for it to be a strong one, DOOMSDAY CLOCK #2 would need to be #1 during its week by a whole lot, while MISTER MIRACLE was #1 during its week just barely. And while that makes sense, I suspect it’s something we’d only figure out in retrospect, after the print data was known.

    I also looked at the reorder charts for December. Neither weekly leaders Walking Dead #174 nor Mister Miracle #5 made the Top 25 reorder list by dollars any week of December. Meanwhile, HAWKMAN FOUND #1 — which has led the reorder list on the comics side of things three weeks running now — made #2 during its first week on Comixology and is 86th at the moment.

    I don’t know what the upshot is, other than digital buyers and print buyers are not necessarily the same group of people (which we already knew) and that the digital list tends to reward buzz more quickly (which makes sense). Tracking them does yield a number of interesting things and it may be worth doing. Drawing more direct connections between the digital and print lists, however, may be a frustrating exercise with inconsistent results.

  10. There’s a ton of ways that this comparison makes no sense, but the biggest is that the “analysis” doesn’t account for variant covers on print books. Publishers have NO IDEA how many readers buy physical comics since a chunk of sales are cover driven.

  11. Interesting analysis and conversation. I’ve always assumed the Comixology bestseller list was similar to an Amazon or New York Times list, where both total sales and recent sales trends propelled books higher on the list. If that were the case, you’d expect new releases to rank highest especially the better selling ones. Who knows, though? Perhaps the different platforms give different rankings to try and identify similar types of readers’ habits?

  12. WRT VENOM INC, my assumption is that CMX isn’t “pushing” the book to all regular ASM readers automatically. With the print copies Marvel/DCD gave VENOM INC as different “series code” than ASM, which took substantial intervention from every retailer, everywhere, to get it ordered/pulled properly.

    -B

  13. This is a very interesting article and comments thread.

    Almost every series referenced is Marvel or DC. However, the two Image digital comics mentioned: Kill or Be Killed (by Todd) and Walking Dead (by John JM) both appear to have overperformed. Watching whether Image overperforms in general would be very interesting.

    Some publishers including Image, IDW, Archie, Oni and Valiant allow DRM free downloads. Others, including Marvel, DC, Dark Horse and Boom don’t. It would be fascinating to analyze whether this has a sizeable impact, minimal, or none at all. I suspect it isn’t a factor for most buyers, but it is actually a very different proposition.

    @JJM – churn would be interesting, but total number of unique Purchasers in a year would be first thing I’d want to know.

    @Egbert475 – why does the lack of variant covers mean this comparison doesn’t make sense? As others have mentioned excluding variants is a better metric for establishing readers.

  14. If one factors out the #1s and variant covers (for higher-numbered issues with variants, by comparing with earlier issues w/o variant), comiXology’s weekly top 10 list actually looks a lot like Diamond’s top 40-50. Same goes for comiXology top 20 and Diamond top 80-100.

    Some of the outliers for Image include WicDiv and Invincible. Saga’s pretty solid at #1 for week of release. Iirc, Saga actually does better than Batman in digital.

    Super Sons #12 is part of a crossover with Superman and Teen Titans hence the unusually high ranking. That said, it’s regularly in comiXology’s top 10 sometimes outperforming Justice League so I’m inclined to think it does have a bigger percentage of digital readers compared to print than the norm. Also of note, when DC had a $0.99 sale on Rebirth issues, early issues of Super Sons ranked high relatively speaking.

    Anyway, looking forward to more articles like this. :)

Comments are closed.