More thoughts on the article

Back on Wednesday (2009-10-21), I got mail from someone I'd run into on mailing lists, prompted by part of my 2009-10-06 entry here about a article. This fellow (Paul Myers, his name is) is an amazing guy, working as a marketer but with a foot also in the geek world and consequently able to say coherent and insightful things about each world to the other—or, at least, to do that to geeks about marketing; it's an assumption on my part that he can do the converse, but it seems likely. (I've met a few people able to bridge worlds like that, all of whom have been very interesting.) He also is a good student of human nature, which I would imagine is a large component of his ability as a marketer.

Anyhow, he thanked me for the pointer (cue xkcd #138) to the article, and remarked that, as resource-hungry as the article was, it was "what people expect in something they perceive as a magazine", and citing an aspect of it as showing they'd thought about the reader's ease of use. This led me to think more about exactly what bothered me. (This is one of the reasons I like corresponding with Paul: sometimes I agree, sometimes I disagree, but he (almost?) always makes me think.)

First of all, to me a magazine is something printed on paper. Trying to duplicate the magazine "experience" in something computer-mediated strikes me as the worst of both worlds, duplicating the problems of a real magazine (such as articles shattered into multiple fragments) while also losing the benefits (paper is still better than computer displays for a lot of things, including bulk reading).

And, tellingly, the thing Paul cited as indicating that they'd recognized the reader's ease of use, a narrow "main column", was something that didn't even exist for me; lynx ignored the layout goop bundled with the content.

The real problem here, to my mind, is that the content provider is not just a content provider but a presentation imposer too. Occasionally this is reasonable; for example, I know an artist for whom the presentation of pictures is as much a part of the art as the pictures themselves, to be taken or left as a whole. But I have this (probably idealistic) point of view that, in general, content is what matters; from that point of view, should have presented the content and let the reader configure the reading tool to present it whatever way enhances that reader's ease of use. That is the Web's potential strength: to use the flexibility that computers make possible to enhance usability for everyone. I'm an outlier in what I find enhances readability, a fairly extreme outlier. That shouldn't matter, because what I find readable should be between me and my local content-presentation tool, not something the content provider should care about. Instead, we have content providers becoming presentation imposers as well, frittering away the protean potential, re-imposing the problems of offline things in their online analogs. (The bloat I noticed is largely just a side effect of this, being mostly the encoding of's particular choice of presentation to impose.) The promise of the Web, the potential, lies in putting flexibility and control in the hands of end users. That isn't what's happening.

Yes, the presentation they impose is fine, for their readership, but that's nearly meaningless, because it's a self-selected sample. People for whom their particular choices impair rather than enhance readability won't join their readership. About all it means is that there are enough people out there who can tolerate their presentation to make up a readership.

Paul writes that it's not about intent to force anything so much as to display a familiar presentation by default. This misses the point: they shouldn't be involved in choosing the presentation at all. That should be a local matter between the user and the display engine. I'm perfectly willing to believe that most users (FWVO "most") like a pseudo-magazine display; fine, let that be one of the available configurations, perhaps even the default, for display software. The problem lies in putting it on the wrong end of the HTTP link. (And, even that aside, calling it a default presentation is at least a bit misleading when there is no way to change that default—shades of "see figure 1". It's actually possible did present some way to change the presentation which I just didn't see; certainly most similar offenders don't.)

Paul also brought up some other points, perhaps most notably ad views. This is a bit of a hot button for me, because the presence of ads at all is one of the main reasons I avoid the mainstream Web. He writes of "additional benefits" in terms of ad views.

Those are benefits for the vendor (the "magazine") and their customers (the advertisers) but, crucially, not the product (the readers). The only benefits I can see to the readers come from subsidizing content, and that is pure fiction, fiction deriving from not looking beyond first-order effects.

Ads don't subsidize content; they actually increase its cost. What they do do is hide a substantial fraction of that cost so it's not easily identifiable. Suppose you have an article that costs $1M to produce and is read by 1M people. We can then think of its real cost as being $1 per reader. But now, let us suppose that it's "subsidized" by ads to the point where its nominal cost is only $0.10 per reader, with the other $0.90 coming from advertisers. Where do the advertisers get that $0.90? Why, they get it from their own customers, of course, with a substantial markup for profit margin, paying for the advertising department (as distinct from paying for the ads themselves), sales tax, etc. Let's suppose this overhead is only fifty percent or so (I suspect two hundred percent is more realistic, but whatever). Then, that $0.90 per reader represents $1.35 coming out of advertisers' customers' pockets. In the aggregate, then, that $1 article is actually costing people $1.45, of which $1 goes to pay for the article itself and the other $0.45 is being leeched off by the various overheads I mentioned a few of above.

Of course, in any particular case the numbers won't be exact. But money isn't being created or destroyed here, just shuffled around; some consumers will pay less, some more, than their fair share, but in the aggregate, the above is what's going on: advertising, far from subsidizing content, actually increases its cost, just diffusing much of it, spreading it out to the point where it's difficult to identify.

It doesn't help that there's a Prisoner's Dilemma operating to force everyone to advertise even though everyone would be better off if nobody did; that's how advertising got started, after all.

On rereading, I find this reads as though I'm being critical of Paul. That's not what I want at all, but it's semi-inevitable because I am critical of what he's describing. I greatly respect Paul; he's intelligent and insightful and is a good deal better than I am at things psychological, especially practical applied psychology. And he provokes me to think, which is one of the greatest services one human can do for another. I don't even think he's wrong in this case; what he writes is true, as far as it goes. I'm really writing about problems in our society—societies—with Paul's remarks being just the provocateur that set me off.