Tuesday, 4 December 2007

Online Advertising's Dirty Little Secret

Woah. So all of sudden the Facebook thing happens and NOW everyone is saying

"Yeah, but do online ads really work?"

A discussion that has been going on in Advertising circles for a long time.

Personally I have always felt that the "trackability" and "measureability" of the Web has been overstated. For a new medium we tend to have pretty old media measurement models with things such as CPM rates (cost per thousand impressions) and click through rates. Even with more "so called" sophisticated targeting and tracking, we still end up asking the same question, do online ads 'work? And what does everyone mean by work anyhow?

Does it mean someone saw them?
Does it mean someone clicked on them?
Does it mean a particular action happened like a pay-per-action model?
Does it mean it generated a lead that led to a sale?
That someone filled out a form?
Does it mean someone was tracked until they purchased something on an e-commerce site?
Was it one sale? Was it a lifetime customer?
Does it mean they like the brand more now? Eh?

What makes matters worse (and I'm not even talking about click through fraud) is that online discussions tend to ignore other forms of media and advertising (because we ALL know advertising doesn't work!). Maybe they clicked on an ad but only did that after their had been exposed to TV, print ads, some radio play as well as a direct mail flyer.

I think what this all calls for is some realistic discussions around measurement, industry wide new standards for ways we measure and broadening out the definition of online advertising from banner and google ads to include other forms of online communications (social media, corporate websites etc.).

Online advertising's dirty little secret is out. Now it's time to do something about it.

(photo credit "shhhh" http://www.flickr.com/photos/vincross/182183355/)


Ben said...

I don't see why people care about clicking so much. If I see a billboard for Cola or Burgers I am not going to click on it. I'd be more likely to feel nauseated than anything.

Mark said...

Of course they work, because if they didn't, why would companies be paying/charging so much for them?

It's just that we don't know quite HOW advertising - online or offline - works. By this I mean, we don't really know, aside from a set of positivist hypotheses that have been proven to demonstrate themselves (big deal), how and why people decide to buy specific stuff.

The reason there are metrics is that people won't pay good money - or even bad money - if there aren't metrics that have a strong correlation to the money being spent. It's even better if the metrics correspond to the "real world" since everyone knows that what happens online "isn't real" (unless it involves someone's income, at which point it becomes real, really fast).

Feh! Not only does the Emperor have no clothes, he lives in a castle of cards.

Leigh said...

banners are pretty much the billboards of the web...

@mark lots of people trying to figure out measurement models. I think that if you can measure the impact of say, a waste treatment plant on a community, you should be able to measure an Ad campaign. The problem is there are so few standards and many advertisers resist performance models bc then they'd be accountable. And really, who wants that? Dang, not even half the clients want that....

Mark said...

Interesting metaphor, Leigh - comparing an ad campaign to a waste treatment plant. That wasn't deliberate, was it? :)

There is a difference, of course. There is a stronger relationship between waste treatment plant and community health, say, than there is between an ad campaign and overall sales, aside from some very special promotions. If you were to try to pose the question about waste treatment and, say, rate of high school graduation in the community, it would be a far more difficult proposition to figure out.

Certainly, you could perform the type of positivist, deterministic investigation that currently is the vogue among marketing folk. And it may even give you a result that clients might even (want to) believe. It will not actually tell you what is going on, because the researcher is not investigating for complexity, and that's the name of the game, in both waste management and waste treatment ;).

Real Time Web Analytics