Wednesday, March 11, 2009

Can Behavior Really Be Predictive?

Google's announcement that it will be releasing two new behavioral targeting products begs the question as to whether we're still holding onto - and whether we should hold onto - the notion that consumer patterns are predictive in nature. It's an interesting move by Google, when you consider the fact, among other things, the search goliath has its own cookieless browser (Chrome), which is optimized for video/rich media delivery but not the more standard ad units the behavioral model currently uses. Perhaps given its sheer size and market capitalization, Google is looking to expand a few different slices of the pie and see what gains the most traction.

But let's first look at the nature of marketing and advertising in general, particularly as social media becomes more of a focal point in outreach and analytics. Advertising has been largely based on the idea that we can somehow calculate meaning. Marketing has been driven by the idea that we can somehow predict behavior. Media have traditionally sat in flux, waiting to blur the lines between them. And this is precisely where meaning gets lost, at least when it comes to "personalizing" ads.

Social media has taught us that "fixed" targets and "fixed" messages no longer have any real impact, and as such, there seems to be some significant value in the behavioral approach. The idea here is to effectively "retarget" or "remarket" a message or story based on user preferences. So I frequent a page, and based on browsing history and other factors relating to purchases, I will be fed ads that personalize or customize the experience to my needs and interests. If my interests change, so do the ads.

What we tend to forget, however, is that these preferences change at an incredible clip and change dynamically. Our sphere of influence - those people that we crowd around, converse and collaborate with -  can also change instantaneously, greatly affecting the environments we visit and how often we visit them. So will that same ad and its related content then travel with me? 

Preference-based environments are great provided that they provide great content, but as we've seen, display content for the most part is sub-par, which is also why CTRs are still abysmal. Further, as we as consumers seek more and more utility in the ad content we engage with, portability is almost always a sticking point. We've already witnessed this struggle with ad-supported content through the likes of platforms such as Joost and Slingbox. Hulu - which seems to have a much firmer grip on this ad model - also faces inherent challenges, especially with the decreasing demand on inventory and a greater demand on free, premium content.

Perhaps one way to reskin this cat is to adopt more of an online WOM (word of mouth) approach. When you look at the dynamic of word of mouth - in which you have the real-time ability to course-correct messaging based on direct consumer intelligence - it organically seems to work in concert with the retargeting concept. So, in theory, you garner feedback as an ad network or 3rd party vendor, redirect that flow of information, and based on mechanisms like meta- or microtagging, you then index and deliver the appropriate ad content. There are also platform providers/networks like TruEffect (not a 3rd party) who are levering consumer intelligence and "delivery visibility" to create a 1-1 relationship between the ad unit (brand) and the consumer.

This does not solve the redistribution problem (such as having affiliate options for placement based on "new" environments the user visits), but it does provide a window into the possibility of looking at ad content as something truly personal, and, potentially offers more options for a consumer within a particular online environment.  Cross-marketing (and cross-pollinating) messages has been one method to compensate for the devaluation of premium and remnant online inventory, but it is a discipline that is often debilitating to brands and cannibalizes the relationship they carry on with their advocacy bases - it is similar to what sites do when they try to upsell the crap out of you before you've even had the chance to make a simple purchase. So, we'll need to look for affiliate solutions that extend outside of an immediate environment, and media companies can explore new opportunities to "redirect and gather", not only ad content but sources of information as well as monitoring intent to purchase.

So is behavior still predictive? Maybe. But perhaps it is more adaptive than anything else. Remember that you can course-correct a message, but not necessarily a conversation - that is simply not up to you ("you" being a brand), it is up to the consumer. In a peripatetic world, committing to conversations versus set ideals seems to be the most logical means of existence.

No comments: