I've lost count of the number of articles I've read over the years declaring the death of online advertising. The reasons cited usually touch at some point on banner blindness, falling click-through-rates (the average CTR having dropped to less than a tenth of 1%), and the uneven distribution of clicks (Comscore's 'Natural Born Clickers' study for example showing that only 8% of Internet users account for 85% of all clicks).
And yet still it grows. The reason, as Ben eloquently puts it, is that "advertising is a game of what you catch, not what you spill". At the rates which advertisers pay for impressions, the results are sufficient enough to still make it worthwhile ("Most advertising never catches anyone. But that doesn't matter, because as in love, marriage, kisses and fishing, all that counts is what you catch").
So (whilst click-through remains a flawed metric in many cases) the game remains one of trying to gain enough attention to keep it that way. The longterm answer to this is highly unlikely (as some seem to think) to be bigger, more intrusive formats. This lesson should have been learnt over a decade ago with the failure of pop-up formats which simply annoyed people to the point of distraction and negatively impacted user experience.
Instead, data has ridden to the rescue of online advertising. We've seen the increasing automation of trading, and targeting becoming an ever more complex activity based often on data collected through third party or proprietary networks. As an illustration of the importance of data now in the ecosystem, tracking app Ghostery shows me (screenshot above - click to enlarge) that when I visit Business Insider (as a random example) I'm tracked by no less than 30 third party sources. Some of these are social platforms that have developed a distributed presence, some are analytics tools, some are third party network vendors. But of-course this isn't unusual. According to tag management company Tagman, there are an average of 14 vendor third party tags on websites. All of these businesses are tracking you in order to gather valuable data with which they might provide more relevant services and advertising, and so make more revenue.
This is, perhaps, not entirely a bad thing. As Dave Winer once said: "Advertising will get more and more targeted until it disappears, because perfectly targeted advertising is just information". Unfortunately we seem to be some way off that. A case in point being the annoyance of poorly executed re-targeting ads (after visiting one retailer's website I've been followed around by their advertising for more than six months and still never bought anything from them). Use of data in this way though, means that the return is still good enough and that online advertising will be around for a good while yet.
But this does mean that display advertising is increasingly a game played at scale. You need scale to make data work because you need to tag as many different people as possible and then have as great a chance as possible to catch them again, potentially at a moment of intent. In this game, the more data you have the better.
This creates a challenge for those who don't have access to sufficient scale of their own - those content producers who rely, at least in part, on revenue from banner impressions served against good quality content and who find that (even if they serve millions of them a month), their inventory will simply never deliver the kind of payback they need. Without a distributed presence, a huge network, or a data play of your own, you become reliant on that provided by third parties, which inevitably dilutes your margins. As media agencies build out proprietary networks and targeting capability, and other data-rich networks enter the game (Facesense anyone?), inventory becomes increasingly commoditised. The real challenge here is that the balance of power has shifted and it will become ever more difficult for content owners to charge premiums based on context and the quality of their editorial environment.