If marketers want more effective online ad campaigns, targeting methodologies need to improve. We all want campaigns that are 100 percent effective -- delivering perfect messages to perfect audiences at the perfect time in the perfect environment. But as we all know, this goal is unrealistic.
Targeting has gotten better over time. But there is still a lot of room for improvement.
Today, audience targeting, which looks at past behavior to determine what a user might be interested in today, is all the rage. But it was never designed to be a stand-alone methodology. Audience targeting came about because there wasn't enough contextually relevant inventory to feed demand.
Let's say I'm a premium publisher or network with lots of content on a variety of subjects, and I sell out my entire inventory on my "health" content. Rather than claiming 100 percent sell-through, I start dropping cookies on my users and targeting ads on other parts of my site to people who had recently visited health content or conducted a health related search, but were now elsewhere. Thus was born "behavioral," a popular form of "audience" targeting. This also took other forms such as the now-ubiquitous "retargeting" where a user who interacted with an advertiser through a visit to their site is now located via cookie and delivered another ad. There's a strong case that most audience targeting is either retargeting or a thin veneer on top of it.
Regardless, it does work, so publishers and networks used it as a great way to turn generic inventory into something closer to contextual. Advertisers had found a great way to broaden a media buy and reach more users.
However, there are challenges. First, it simply can't scale to desired volumes. Second, there are some pretty serious brand safety concerns. Couple that with increasing consumer privacy concerns, and the decreasing accuracy of being able to peg a cookie to a user and you have an industry that will have to evolve or get left behind.
As it relates to scale, when you're out trying to buy impressions associated with a cookie list, it's tough to get volumes that would be the most effective for advertisers. These campaigns can be effective, but they leave marketing dollars on the table.
When you add in brand safety concerns, audience targeting begins to get a little shaky. Most audience targeting providers simply don't have the technology to ensure 100 percent brand safety. I'm certain that Mercedes-Benz would not find this to be an acceptable ad placement.
Internet users are also becoming savvier about online privacy, with 40 percent regularly deleting cookies, according to Consumer Reports. Moreover, a recent TRUSTe study indicated that only 11 percent of consumers are comfortable with online behavioral tracking, a statistic that does not bode well for the future of audience targeting. In addition, certain categories of advertisers, including health care, finance, and education, are particularly concerned about privacy. Retargeting, in particular, is a big no-no for many prescription drug companies.
Another key factor is that web consumption at home is increasing. Unfortunately, audience targeting doesn't tell advertisers if it's me, my wife, or my 12-year-old using the computer. All they see is the cookie.
Finally, most of the browser providers have implemented "do not follow" as an installation feature, which will further erode audience targeting.
Marketers want to reach the right audience at the right time, and most want to do it in a relevant, brand-safe environment. Despite the challenges I've already mentioned, audience targeting does have its place. But when you peel back the onion, it's really a mixed bag in terms of performance.
I'd grade the industry's current ability to find the right audience at a B+. Some media players have developed good models and technology to deliver on the promise, but many of them are doing spray-and-pray retargeting to get credit for the conversions. Even the good ones can't address the fundamental issues of shared browser and cookie-churn.
How about connecting with people at the right time? Almost by definition, audience targeting is targeting a user after she has moved past the session or search in which her intent or affinity for a particular product was detected. Is she still in the "funnel" or has she left it? Even if she is, is the ad being shown at the most relevant time? She might be at work doing work-related research and doesn't have the time or the inclination to chase the vacation package the ad is offering her. Or worse yet, she already made her purchase, yet keeps getting ads. I'll bet every one of you has been retargeted for a purchase you already made. So, the ability of advertisers to get the timing right gets a B grade at best.
And when it comes to a brand-safe, quality environment, audience targeting gets a solid F. My company constantly looks for and filters sensitive pages, and it's unbelievable how often we see ads from top brands on sites with questionable to downright unacceptable content.
The bottom line is that audience targeting has a clear role to play, but it can't stand on its own.
Audience targeting, including behavioral and retargeting, will increasingly get overlaid with contextual targeting to ensure brand safety and relevance. Smart advertisers who want to reach target audiences in real-time, consuming content that best expresses their intent, will overlay this strong contextual signal with the demographic or psychographic characteristics of a particular site's audience.
As these technologies continue to mature, the goal of reaching the right users in a contextually relevant mindset at the time the ad is being served looks increasingly achievable. If you can do that without raising users' privacy concerns, you're golden. I won't be surprised to see more and more advertisers asking for this hybrid ad model, and industry providers stepping up to serve them.
John Mracek is the CEO of NetSeer.
On Twitter? Follow iMedia at @iMediaTweet.
iMedia Communications, Inc. All Rights Reserved.