Saturday, May 28, 2016

never mind the bull elks

Bull Elks face, what the economist Robert Frank calls, a collective action problem.

Over time the genes for bigger antlers have prevailed because larger antlers make an individual bull more likely to win in a fight with a rival bull.

Bulls that win more fights have greater status in the herd, and access to more lady elk action etc, so there’s significant benefit on an individual level.

However, large antlers force on their bearers some less than optimal compromises.

Getting about in the elks natural habitat - dense wooded areas – is hampered by giant antler-age, and therefore increases the risk killed and eaten by predators such as wolves.

As a group, elks would be much better off if each bull's antlers were much smaller, bulls with smaller antlers would be better able to escape predators.

But as long as there are some bulls with the big antlers the less well endowed will always come off worse in battles, and therefore less likely to pass the genes for those smaller antlers into the next generation.

The individual payoff on larger antlers is thus substantially larger than the collective payoff.

The collective action problem (on behalf of bull elk genes) is that even though it would be better for all bull elk if everyone’s antlers were smaller, it’s not in any individual bull's interest.

Any trait that emerges because it helps individuals compete in battles against members of the same species runs the risk of resulting in a handicap for the species as a whole.

This might be what's going on in online advertising.

While publishers at one point may have shared common interests with advertisers - reaching readers -each also had conflicting interests, principally around how each made a buck.

Enter stage right, 3rd party adtech with the promise of identifying and tracking prospective customers on behalf of advertisers, thus devolving the publisher from their part of the conundrum - providing the audience - and 'incentivising' huge volumes of fraudulent traffic that drives down the value of CPMs.

This created a wedge - adtech are not interested in the needs of readers, but publishers became distanced from the needs of their readers and their advertisers - and before you know it we’re in the whole mess of aforementioned fraud, malware and goodness knows what else.

There's an argument for banners as simply low-cost mini-billboards, a super cheap vehicle for impressions. So cheap that the fact that most people don’t see them, and that a significant chunk of views are fraudlent doesn't really matter. The market has adjusted the cost down to make them good buy even factoring in the waste (this is proper wastage, not 'costly' waste that carries signaling power kind of waste).

Meanwhile, the advertisers who really only want to efficiently reach punters, start scratching their heads.

The individual payoff for online publishers on adtech antlers once promised to be large, now the collective payoff is more like being stuck in the woods with a wolf.

Don Marti wrapped this up nicely with 3 points.

When bottom feeders pay indiscriminately for fraudulent and legit banners, they are funding malware development. 

The glut of ow-quality impressions don’t pay for original sites, and fuel all the other bad behavior.

When advertisers buy on markets that include both legit and fraudulent inventory, publishers pay for fraud in the form of lower CPMs.


It's a collective action problem.