Commentary

'You Guys Should Talk'

Randall Rothenberg of the Interactive Advertising Bureau, meet Jim Sterne of the Web Analytics Association. There are a hundred reasons why you two should be at the same table.

Jim, as you may know, Randall is the new President and CEO of the IAB. Randall, Jim is the head of Target Marketing but also runs an association called the Web Analytics Association. In a previous Metrics Insider, Josh Chasin pointed out that your two organizations are both out there setting standards, a laudable effort.

Yet, the standards your organizations are setting, while they use many of the same terms, are different. Each organization has its own definition for things like clicks and click-throughs. (IAB actually has a technical definition while the WAA basically describes the action).

There is a serious underlying issue here, and it is called attribution. Not only do we have to get the two standards bodies on the same page for definitions, you both need to agree what to do with items like pixels and attribution.

Now, I am not a researcher. Nor am I a Web analytics person. But I see the results of counting by a third party ad server (3PAS) where we know how many people viewed an ad and went to the site through a click-through or a view-through (post impression). The site-side analytics don't understand view-through, and the IAB has not even got it on its agenda for definition, near as I can tell.

Yet, over half the visitors who saw a client's ad do not come through a click-through but visit the site later through the URL line. We can tell this as we cookie them when they are exposed to the ad. Now, I am not going to get into a discussion here as to the veracity of view-throughs, as that is a separate article. We know they are valid at least to some degree and can prove it and will.

But one thing is clear. It does us no good when, as a result of these different standards, the Webmaster says we got 50 sales and the 3PAS system says we got 100. Even if view-throughs are not counted, the differences are significant. Yes Jim, I know that if we deploy two of your members' systems, they end up with different numbers, too. That's another story and one that all need to work on. But it does no good at all when there are standards bodies, trying to define the same thing, that are not at the same table.

So, I am going to make this easy. You are both on LinkedIn. You are both linked to me. Either one of you can send an invite to the other to link, or I will gladly forward it.

I, and probably many others in the industry, am willing and able to join the two of you around the table. But we need to start the discussion.


Next story loading loading..