The microeconomics of the link economy

Some two weeks ago, internet investor Chris Dixon wrote a post in which he challenged some of the conventional wisdom surrounding the link economy. I then wrote a critique of his post, and then in the comments he critiqued my critique. After some digesting, here is a new attempt at the whole story in distilled form. Comments welcome.

1. In the battle between publishers and aggregators, aggregators have the upper hand

To see this, consider the relationship between:

• An aggregator, by which I mean any website that makes its money largely by providing links to content hosted on publishers' sites. The obvious examples are Google News, the Drudge Report and (to a lesser extent) the Huffington Post. But t could also be Twitter - it doesn't matter if the links are authored by the site owner or by its users.

• A publisher, by which I mean a site that hosts content that it "owns". This could be a newspaper's website or any site with a right to host and monetize content. It doesn't matter if the site's owners are the creators of the content or have simply paid for the right to host it, like Yahoo News.

Assume that if either the aggregator or the publisher decides, unilaterally, that the aggregator should not link to the publisher, the linking will stop immediately. (Of course it's easier for the aggregator than for the publisher to accomplish this, but for the argument's sake assume that the publisher also can, whether through legal or technical means).

The key thing to note here is that - under reasonable assumptions - the aggregator's financial loss would be far smaller than the publisher's, typically by an order of magnitude. I won't bore you with the detailed reasons for this (it is standard stuff), but two key assumptions are that 

• If the aggregator were could no longer link to the publisher for free, it would readily find another similar publisher willing to accept the traffic for free

• As a result of this, very few of the aggregator's users would care if it stopped linking to the publisher

This means that the aggregator has all the bargaining power: it can walk away and suffer a fraction of the pain that the publisher would. Because of this, publishers have no hope of extracting payments from aggregators. 

Note, importantly, that this has nothing to do with how big the aggregator is - i.e. it doesn't matter if it is Google News or a niche link-blogger like Martin Stabe.

2. Individual aggregators may be able to charge individual publishers (not the other way around)

At this point you could say: "The aggregator is profiting from the publisher. If the aggregator doesn't pay the publisher, it is free-riding. This is unfair."

One way of dealing with the question of fairness is in terms of how the "pie" of value jointly created by the publisher and the aggregator is divided. Take the profits made by the aggregator thanks to the publisher, the profits made by the publisher thanks to the aggregator, add the two, and you have the pie. A "fair" deal could be to split the pie in half, so that the party making more profits makes a payment to the other to compensate.

Above I said that any given publisher stands to lose a lot more than any given aggregator from a stop in linking between the two. But a different way of saying this is that the publisher has a lot more to gain from accepting the links than an aggregator has by making them - that is, the aggregator is keeping a far larger share of the pie. The aggregator could stop linking to the publisher, suffer a minimal loss, and then go the publisher and propose to resume linking if the publisher would only agree to give half of the upside to the aggregator. A rational publisher would agree. 

The fact that this is not happening suggests the convention of "free" is stopping aggregators from playing this card. "Free", publishers' dreaded enemy, is their friend here.

3. But aggregators may be free-riding on the publishing industry

My claim above that publishers have a lot more to lose than aggregators from a halt in linking relies on the assumption that aggregators can readily replace any publisher by another who is willing to accept traffic for free - i.e. that the publishing industry is competitive and that publishers don't talk to each other. But what if this were not so? What if publishers could (legally) team up in a cartel?

If all the publishers were to stop all traffic coming from a given aggregator, in this case it is not true that their loss would be far greater than the aggregator's. Whose loss would be greater depends on whether the aggregator is bringing traffic that the publishers would not otherwise have. More specifically:

• If all the traffic is incremental then publishers may have more to lose than aggregators, because they can probably monetise this traffic better (e.g. more pages per visitor, higher CPMs, etc). Value is being created, and under a no-payment scheme publishers would still be getting more than half the pie

• If none of the traffic is incremental then publishers have nothing to lose from blocking the aggregator. They could drive a hard bargain and demand that the aggregator pass on half its revenues - or more

Reality is likely to be a mix of these two scenarios. If, say, half of the traffic that publishers get from aggregators is genuinely incremental to what they would get without them, this would mean that (a) value is genuinely being created (i.e. people read more news than in a world without aggregators), but (b) under today's no-payment scheme aggregators as an industry are capturing more of this value than publishers as an industry.

Unfair? Perhaps, but then business has never been fair in that sense.

Comments

[twitter.com/cdixon]

I think we bascially are in total agreement. I just assumed a cartel would be illegal so didn't bring it up, hence my thought experiment imagining that the ratios were reverse - aggregators / search engines were plentiful and interchangable, publishers concentrated and scarce.