Friday, September 14, 2007

Figure Skating

Heidi MacDonald posted a few pieces on comics sales charts at The Beat last week (One, Two, Three) that spawned quite a bit of discussion in their comment sections, with input from industry observers Dirk Deppey, Eric Reynolds, Paul O'Brien and Tom Spurgeon, among others.

As it frequently does in this context, the question of the usefulness of the available direct market sales data comes up. And not surprisingly, not everyone's convinced of it - the most vehement criticisms, in this instance, come from creator Brian Wood (DC Comics/Vertigo's DMZ, Oni Press' Local). Wood is arguing, in a nutshell, that the direct market sales index information provided by Diamond Comic Distributors, the sales estimates calculated from it by and others, as well as the frequent publication and analysis thereof, are wrong, harmful and - that's the impression - generally and wholesomely evil.

Now, don't get me wrong. I have all the respect in the world for Wood's work, his experience and his contributions to the medium and the market. And given that he makes a living creating comics, I can certainly see where he's coming from, as far as his attitude towards the public discussion of sales figures is concerned. That said, though, Wood's objection basically seems to be that he doesn't like what the available sales data says, and therefore everybody should stop reporting it and talking about it. Naturally, I don't think that's a very useful approach - or a very realistic one, for that matter. But let's look at Wood's points.

For starters, Wood makes a comment on reorders that seems rather fuzzy.
The life of a book does not begin and end with its initial orders, and even those IVC2 [sic] charts that claim to also include reorders, they only report some reorders and for inconsistent periods of time. Again, incorrect information.
Well, first up, it's not who put together the charts and decide what's included and why, but Diamond Comic Distributors. merely use the Diamond charts and index information to calculate sales estimates of those books included in the charts. (And, by the way, I agree that Diamond's notion of using a given month's issue of Batman as the "100%" benchmark for their index points is fairly asinine.)

Second, and more significant, the charts do include reorders. However, they're obviously the "Top 300" charts (or "Top 100," in the case of the "Graphic Novel" chart), so naturally they only include those reorders which actually register on these charts.

Which can happen in two ways. If the reorders ship in the initial month of a book's release, they're all counted and added to the book's initial sales. Let's take the July chart as an example: Wood's DMZ #21 was released in the second week of July. What this means, in terms of reorders, is that any and all copies of DMZ #21 reaching stores in the third and fourth weeks of July are added to those initial sales in the first week, as far as the chart is concerned.

If the reorder copies don't reach stores in the month of the book's initial release, on the other hand, the only way for them to appear on the chart is if there are enough of them in any given subsequent month to make the chart. Again using the July chart as an example, an estimated 2,354 units of World War Hulk: X-Men #1 appear in slot No. 291. The issue in question was first released back in June, so those are reorders, and they're evidently strong enough to place the book on the chart again in July. If the sales of reorders for World War Hulk: X-Men #1 had been fewer than 2,000 in July, it likely wouldn't have appeared on the chart, since the lowest-selling book on the Top 300 chart that month sold an estimated 2,086 units.

Obviously - and this is where I agree with Wood - this approach isn't ideal. It means that reorders which appear after a book's initial month of release but aren't strong enough to make the chart on their own are inevitably going to fall through the cracks. But while that's most definitely a limitation, it's arguably a minor one in most cases, because the same demand that keeps generating reorders for a title tends to result in additional initial orders for subsequent issues: If you're a retailer and constantly have to reorder copies of a given periodical without ever reacting to that by adjusting your initial orders for the next issue, you're doing something wrong. So while not all reorders may be tracked by the Diamond charts, the accuracy of the trends they show shouldn't be affected by that.

And, really, while the Diamond charts may not track some reorder activity, they're not hiding that fact, and there's nothing inconsistent or inaccurate about the information they do provide. The information is simply not as complete as we'd like it to be, in an ideal world. I believe it comes close, though, in most cases.

In another comment, Wood criticizes the term "actual sales" in the analysis of Diamond's periodical sales charts.
There is nothing “actual” about them at all. Still mostly North American pre-orders only.
Well, that's incorrect. What the Diamond charts "mostly" report is indeed sales to North American retailers, and not just preorders. As I've explained above, some reorders regularly slip under the radar of these charts, but that doesn't change the fact that they report sales - including most reorders, most of the time.

And, by the way, yes, I'm aware that's not the same as sales to customers. Information on actual sell-through would be very welcome to anyone with a faint interest in the overall market, I'm sure, for a variety of reasons. But in a market in which 99% of the product is non-returnable, it's simply not required when it comes to the question how well publishers and individual books are faring commercially in a given month, which is, after all, precisely what those charts are meant to track. Blaming the available sales charts for the system they provide data on, surely, is putting the cart in front of the horse.

I agree with Wood in so far as the term "actual" can be misleading in this context, though, because of all the qualifiers you need to understand the charts. But I don't think that mischief on anyone's part is involved. Rather, the term is owed to the history of the charts. Several years back, you see, Diamond did provide data on preorders only. That changed in March 2003, however, when they made the switch to reporting the sales of books as they actually shipped. To highlight the difference between the two approaches, the term "actual sales" was introduced. What it actually means, of course, is "actual sales through Diamond Comic Distributors to comics retailers in the North American direct market, not including reorders which shipped after a book's initial month of release and failed to make the chart again." There you go.

That about covers Wood's objections to the systematic limitations of the available sales charts. They're fair game, really: The limitations exist, and for the sake of transparency and full disclosure, they can't be pointed out often enough. Where we part ways, though, is in the conclusions he draws.

Wood says:
[A]nalysis of an incomplete picture can be a harmful thing. Consumers and retailers can see a monthly title dropping low in these charts, read some snarky one-line analysis about how bad its looking for a book, and maybe decide its not worth buying/ordering any more if its likely to be canceled. When the reality of the situation can be very different.
Later, he adds:
Seriously, everyone, the name of the game is perception, and most readers aren’t going to scour an article or a comments section looking for the grand big picture or the little caveats sprinkled around. They’ll breeze through, reading the title, looking at the charts, and walk away depressed that their favorite book is tanking, or whatever, never knowing they’re only getting part of the story.
I've had this discussion a few times in the past - a couple of times with Wood himself, I think -, and these kinds of statements never fail to annoy me in their glorious wrong-headedness. What the Diamond charts provide, at the end of the day, is information - nothing more, and nothing less. Is it incomplete information? Of course it is. But which publicly available sales data on pretty much anything can claim to be "complete"? It's not like Diamond are trying to hide the limitations of their charts, either - in fact, they've always been relatively upfront about them. And the same goes for most of the articles that frequently publish, process and interpret the information provided by Diamond.

Bearing this in mind, the gloomy image of hordes of customers and retailers, lemming-like in their sudden divorce from reason and experience, dropping supposedly low-selling titles left and right after "breezing through" a sales report does not seem to be a very likely scenario to me. Nor, by the way, do Wood's comments suggest a very flattering notion of his readers and their wits. And even if all those vague speculations were justified, for that matter, since when is the mere possibility of people misunderstanding information offered to them considered to be a valid reason against its dissemination?

As long as the origins and limitations of the available sales data are made clear in a given article, I don't see any convincing reason to view it as inherently "harmful" to anyone or anything. Rather, I regard it as a valuable, sober counterpoint to the publishers' own frenetic hype machines, which makes the direct market and its workings a little more transparent to everyone. And, honestly, if the sales reports help retailers - who, lest we forget, probably have their own experience and cycle sheets to draw upon - to get a better idea of how well a book is doing across the board, then surely that's a good thing in the long run.


brian said...

"Wood's objection basically seems to be that he doesn't like what the available sales data says, and therefore everybody should stop reporting it and talking about it."

Fundamentally missing my point again, Marc. I simply have stated, repeatedly, that these numbers simply need to be labeled accurately. Not gonna hash it out again, as I'm on record as saying this in a number of places.


David said...

Your thought seems reasonable until one point, where I would at least diverge:

"As long as the origins and limitations of the available sales data are made clear in a given article, I don't see any convincing reason to view it as inherently "harmful" to anyone or anything. Rather, I regard it as a valuable, sober counterpoint to the publishers' own frenetic hype machines, which makes the direct market and its workings a little more transparent to everyone."

It seems to me that your feelings on this depend largely on whether you see the analysis of Diamond data as independent from or intertwined with the hype machines. (I'm not so much talking about the data itself as the analysis of it; claims like "World War Hulk leads Marvel to dominate the market," which seem to follow publication of the data like leaves from a tree, and which seem to me intertwined with the PR machines.

In that sense, I do think they can have a negative effect; retailers and customers do have their own experience to draw upon, but the gluts/crashes of the 1980s and 1990s show us that hype machines can distort.

That said, lucid and clear analysis, and an engaging read (discovered through the Journalista blog).