Proceeds from the ads below will be donated to the Bob Wuesthoff scholarship fund.

Wednesday, February 07, 2007

OT: In College? Nielsen Says You Didn't Count — Until Last Week

One of my ongoing bugaboos here is the apparent incompetence with which multimillion-dollar decisions are made in the broadcast business thanks to the likes of Nielsen and Media Metrix, who have trouble doing their jobs on the Internet. Most recently, it was a topic that arose because of the Times fiery memo straightening out the alleged fuddy-duddies writing the news at Spring Street; much handwringing went on due to the supposed declines in site popularity shown by the non-scientific Alexa toolbar, and by the purportedly authoritative Media Metrix.

The topic also arose during the recent Church vs. Super Bowl dustup, in which the excrement ran downhill from Nielsen (which wants to keep people at home so it doesn't screw up their statistical models of what constitutes a "household" and thus ultimate viewership) to the NFL and thence to the beatdown at Indianapolis' Fall Creek Baptist Church. The supposition — one I've made, as well — is that Nielsen ought to at least have a half a clue about measuring TV audiences because they tend to be larger than Internet audiences. Nonetheless, criticism of their methods of even that task is ancient, dating back at least to the TV diary (long abandoned in favor of set-top snoops).

But now comes word that Nielsen hadn't been doing their job so well in measuring college students, and for the first time, college student viewing habits in dorms are being recorded. As a result, "Grey's Anatomy" and "Ugly Betty" recorded record-high ratings, and other shows, such as "Scrubs", saw significant boosts in their numbers.

Advertising on the Internet has made it possible, for the first time, to have exquisitely detailed information on how people react to advertisements. The first phase of this was for customers to act on the old complaint, "I know I'm spending too much by half, but I don't know which half." Armed with response and subsequent action data (i.e., maybe some users clicked through to the target site, but did they buy once there?), advertisers can filter those things that work, and with the immense inventory websites permit, to cut their costs to the bare minimum. As television (and media generally) becomes more and more like the Internet (smaller, flightier audiences with more channel choices), advertisers will demand better quality reporting — just like they get from the Internet.

Interestingly, I knew of a couple of houses in college that were actually serving as Nielsen "families." They were just collections of 8-10 housemates living together (off campus) for a couple of years.
I'm not so sure that Internet statistics are all that accurate, either. Sure, you can measure page views and click-throughs accurately, but that doesn't mean that the ads were actually looked at, or that the click-through was purposeful.

I'd be willing to bet that at least 75% of all ad click-throughs are accidental. I'm pretty sure at least 95% of the time I click on an ad, I didn't mean to.

There's still a ton of room for improvement in making advertising more cost-effective.
I had a Nielsen box in my on-campus college apartment one summer at Stanford. This was back in the early 90s. So I claim "first".
Oh, to be sure, Ken, even when you do have the means to do direct measurement, it's not always telling you something as important as you might think. But that's what I mean when I say that page views aren't necessarily interesting in and of themselves; the belief that it was became the fatal flaw in many early dot-com era companies. With the means to discover value on their own, advertising customers rapidly realized that untargeted web CPM was largely a crock. I don't know if I'd say that 75% of all aggregated ad clickthrus on the Net are accidental, but I would also suggest that could be as high as 100% for isolated cases if you're dealing with something like a typosquatter (and there are a bunch of guys who make a bunch of money from those sites). It could also be much lower if you have the means to put very targeted ads in front of your customers, as Amazon is doing with their suggestions-based-on-past-browsing links. This is something even Doubleclick doesn't have the ability to assemble yet, or at least, they didn't last time I looked.

Post a Comment

Newer›  ‹Older
This page is powered by Blogger. Isn't yours?

WWW 6-4-2