In a recent commentary published on MediaPost, George Simpson of George Simpson Communications, talks about the follies of behavioral targeting and other web-based analytics tools that seem designed to tell us everything, but often fall short of providing advertisers or web publishers with an accurate description of their audience's motives. Simpson points out that while frustration abounds, it is a little short-sighted to dismiss all analytics as the work of snake oil salesmen and keep in mind that the benefits of what we've learned through these tools far outweighs the negatives of what we haven't.
And I have to say, I agree. But there is one other piece of the puzzle I believe we should look at and that is expectations.
Expectations have changed for your average Internet user as online advertising and behavioral targeting have matured. We no longer accept that our three-year-old playing a Thomas the Tank Engine game might be served an ad with PG or PG-13 rated content. We have learned to filter out repeated ad content (aka the ads for credit scores, online education and hula girls with swaying hips). Our eyes are now drawn to much more interesting and targeted content: an ad for our favorite shampoo, an upcoming TV show, or the latest blockbuster. As consumers, we feel a certain sense of entitlement that when we visit our favorite websites, we won't be confronted with "junk" on the page, but rather related advertising content that doesn't detract from our experience.
This is the expectation web publishers have as well. While there are still the $0.50 CPMs that allow ads to be served on the most random of pages in order to bring down the eCPM of an overall ad buy, web publishers would prefer to serve their constituency ads that matter to them. Publishers feel (and rightfully so, in my opinion) that crappy ads reflect poorly on their site. It is of course a fine balancing act between revenue and reputation, but it appears that all other things being equal, publishers would rather serve premium and/or quality advertising as opposed to not.
So then, whose expectations are skewed? It appears that consumers and web publishers, the absorber and pushers of the information are just fine with using collected behavioral data to enhance their online experience. Who doesn't get it?
Now, look, I love advertisers and agencies as much as the next gal, but there is still a disconnect between what is actually feasible to measure and what advertisers want to see measured. As with all advertising mediums, advertisers want guarantees. They don't want to throw darts at a board (or money down the toilet); they want to know that their ads will be seen by qualified consumers who will actually take action.
Um, we all know that's not possible right? I can no more guarantee that a visitor to the Boston Globe's sports page is going to click on an ad for Sports Memorabilia, than I can promise that a visitor to the New York Times Arts & Entertainment page will click on an ad to win a trip to London's West End. Now, based on positioning and target, I would say it's highly likely these visitors might do both of these actions, but promise? Guarantee? Nope, sorry. All too often, advertisers demand information that we in the industry just can't get. However, because our expectations have become skewed and because the advent of the Internet has made the line between public and private very fuzzy, the perception is that we should be able to know what site a 35-year-old man from Cleveland, Ohio visits after watching the latest episode of "The Glades" and what content he looks for and what advertising he takes action on.
I can get you Cleveland, I can get you 35 (probably more like 29-35), I can get you male, I can even get you time of day, but I can't promise that the guy who looks up your ad just finished watching a particular episode of a particular show. I also can't promise that he liked it.
In my previous roles with ad agencies, it always startled me how much advertisers wanted these types of guarantees. I understand the need to know that your money is being spent wisely, but there are so many other variables that lead to someone taking action on an ad: mood, alertness, financial situation, etc. And trust me, if I could accurately predict clicks, I'd be richer than Trump.
All advertising: print, radio, digital--is bought on the predication of best guesses. Through study, surveying and testing we believe we have a fairly accurate picture of who reads what types of magazines, who listens to what types of radio stations, who visits what types of websites. But I doubt we will ever get to the point where we know all of these things for certain (unless Big Brother gets a whole lot bigger, and maybe puts a chip in our head). And, until we admit that the tools we use (i.e. Nielsen and comScore) are not all that accurate, the data we extrapolate from these sources will be a best guess--a really good guess, but still a guess.
Now, obviously there are less than reputable agencies out there that sugarcoat and hide actual web stats to convince advertisers to waste money placing their ads where they truly have no hope of getting any traction. But I like to believe that more often than not, we are all doing our best to steer our clients in the right direction. It takes as much of a leap of faith for us as it does for them.
And until the answer: Based on all the research and evidence we have, we think this is going to work, is acceptable, we'll all just be making promises we're not sure we can keep.