Clay Shirky, paywalls, heavy sigh

I like Clay Shirky. We have met a couple of times. We are on the same page for lots of issues. And I really admire his way to economically express some important issues. But recently I have become concerned that his arguments are a little too snappy and don’t really stand up when pushed. That said, I am sure that in the spirit of communicative Internet, he would appreciate being called on where he is weak.

The most recent example was his post yesterday on newspapers and paywalls that began, “[t]his may be the year where newspapers finally drop the idea of treating all news as a product, and all readers as customers.” Sure, OK, I’ll bite. What does that mean?

The starting point is that paywalls have failed. The strong paywalls (e.g., the Times in London) have lost so many online readers that this is now dwarfed by offline readers. For the “payfences” (as Tom Standage calls them) — the NYT being the prime example — there is a mixed message:

The article threshold has often been discussed as if it was simply a new method of getting readers to pay, to which the reply has to be “Yes, except for most of them.” Calling article thresholds a “leaky” or “porous” paywall understates the enormity of the change; the metaphor of a leak suggests a mostly intact container that lets out a minority of its contents, but a paper that shares even two pages a month frees a majority of users from any fee at all. By the time the threshold is at 20 pages (a number fast becoming customary) a paper has given up on even trying to charge between 85% and 95% of its readers, and it will only convince a minority of that minority to pay.

He goes on: “This is the commercial equivalent of the National Public Radio model, where sponsors reach all listeners, but direct suport only comes from donors.” So far so good but this doesn’t quite get us to failure. If the Times wasn’t making money from online advertising, they have lost nothing. If the NYT‘s readers were not visiting too often anyway, then they are just price discriminating and charging those who do. What is more they are gathering information on all of them and they can use that to sell advertisers different products. It isn’t clear that it is commercially bad from that alone.

However, what it does do is cause newspapers to focus on what their paying customers want. Shirky then loses me:

Paywalls held out the possibility, however illusory, that if all readers could be treated as customers, the organization wouldn’t have to pay much attention to them, except in aggregate. Threshold charges blow that up; a single fee-paying user will generate hundreds of times the revenue of the median, ad-viewing reader. This subjects the logic of the print bundle — a bit of everything for everybody, slathered with ads — to two new questions: What do our most committed users want? And what will turn our most frequent readers into committed users? Here are some things that won’t: More ads. More gossip. More syndicated copy. This is new territory for mainstream papers, who have always had head count rather than engagement as their principal business metric.

Wait a second. There are two parts here. First, that customers will sort into different types based on whether they are transient or intensive visitors to an outlet. That is true. Second, that the sorting will be with ‘serious readers’ being intensive and ‘non-serious ones’ being not. He elaborates on that:

Celebrities behaving badly always drive page-views through the roof, but those readers will be anything but committed. Meanwhile, the people who hit the threshold and then hand over money are, almost by definition, people who regard the paper not just as an occasional source of interesting articles, but as an essential institution, one whose continued existence is vital no matter what today’s offerings are.

Is that so? It is certainly far from obvious. I could re-write that passage plausibly as:

Great in-depth stories will always drive page-views through the roof, but those readers will be anything but committed. Meanwhile, the people who hit the threshold ad then hand over money are, almost by definition, people who regard the paper not just as an occasional source of interesting articles, but as an essential part of those days, one whose continued existence affords them 25 minutes of enjoyment regardless of what today’s offerings are.

This statement is wrong — well at least without any empirical proof — but it as equally wrong and equally plausible to Shirky’s. And it matters because which one is true will drive how the newspaper positions itself to appeal to paying customers. So again, we surely can’t conclude that “[t]he most dramatic change, though, is that the paying users are almost certain to be more political, and more partisan, than the median reader.” That may be true for some papers but really, the whole industry?

So my point is that if Shirky thinks that newspapers are going to reward loyalty from customers who want great journalism, independence and other nice things, then it had better be the case that those are the paying customers.

But more so, there is some importance to the notion of what people are paying for. For Shirky, readers are paying potentially “to give something back.” But he starts his post with this great quote from Paul Graham:

Consumers never really were paying for content, and publishers weren’t really selling it either…Almost every form of publishing has been organized as if the medium was what they were selling, and the content was irrelevant.

And then Shirky doesn’t return to it. But, in my mind, this gives us a way of making sense of the NYT payfence without appealing to “gift giving non-customers.” The NYT is selling a medium. The medium is — if you want us to curate a large portion of your daily news and then present it to you in a convenient — cover to cover — format, pay. That is what the iPad app does and it is what the print newspaper does. It is not what the NYT website does. In that respect, the NYT, to keep them paying, has to treat them as customers of curation and to do that job well. That is, some consumers still want the bundle. And believe it or not, that will rely as much on technologies for personalisation as it will on traditional top-down functions. Do that well and the content will find the customer.