ComScore recently released an interesting White Paper titled How Online Advertising Works: Whither The Click which presents ample evidence that using clicks to measure the effectiveness of online display advertising is a bad idea. In particular, the paper cites Starcom research that suggests no correlation between display ad clicks and brand metrics. Good thing, too, as ComScore also reports that average click rates on display ads in 2008 were less than 1%.
Looking at alternatives to clicks, ComScore found online display ads do in fact increase:
- Visitation to an advertiser’s web site
- The likelihood of consumers conducting a search query using an advertiser’s branded terms, and, most importantly
- The likelihood of buying the advertised brand (either online or at retail)
These findings echo arguments that agencies and display ad service providers such as Microsoft have been making for quite a while regarding the issue of conversion attribution, but I don’t wish to go there again. Instead, let’s go back to the conversation about advertising on Social Networks, which the New York Times recently weighed in on.
The ComScore report offers useful breakdowns by advertiser sector and activity metric, but doesn’t speak to many critical variables, such as targeting method, ad format, or the context of ad placement, all of which can have strong influence on the pricing, if not effectiveness, of an ad. The issue of pricing won’t go away: there is little doubt that economics are driving advertisers toward performance pricing, and ComScore data supports this by noting that Nielsen data shows a 6% decline in online display CPM-based revenue in H108, while Competitive Media Reporting shows overall display revenue (CPM and CPC/CPA) rising 8% over the same period.
All of this suggests that, if social networks are to turn the corner on advertising, they’re going to have to jump into the breech of engagement metrics and come up with CPA pricing methods that track and monetize events that they can prove correlate with the actions above.
Clearly this is a tall order. Among the problems are how to correlate a valued action with an earlier ad exposure without relying on shaky technologies like cookies. This is one of the reasons why Google Friend Connect vs. Facebook Connect is such a significant battleground: among the advantages their use will convey is the ability for social networks to track web site visits and even tie them to click-less CPA actions.
The other side of the challenge is for social networks to get much better at understanding and targeting the context of advertising. The advantage they have, so far apparently underutilized, is the ability to understand not just the context of a web page, such as Google uses to target AdSense, but the context of a community, so they can gauge each ad’s relevance, and hence predict – and price by – some meaningful delayed response. This means potentially sacrificing near-term revenue by actively reducing irrelevant advertising.
Google, the pioneer of CPC advertising, appears poised to move beyond clicks by using its formidable infrastructure to take on more of the risk that separates CPM publisher aspirations from CPA advertiser aspirations. On the other side are Facebook and many other web publishers who would prefer to sell their own inventory, and agencies who need to maintain control of high-end media planning. In all cases, we need to move beyond a fixation on clicks.