Alternative Metrics: The Session

Two weeks ago I started this discussion about a search for standards and alternatives to the existing forms of online media currency. Last week I offered thoughts on “audience,” or “rate base,” as one alternative to the currently accepted “atom” of the online advertising universe, the impression. But I’m not the only one talking about this stuff. Over the last couple of months, and particularly in the last month, the discussion about standards and alternatives to the way online media is purchased has found new life.

In these very “pages” of ClickZ, there has been good coverage of the topic. The two most constantly interesting voices on this topic have been Tom Hespos, who wrote about defining ad views, and Tig Tillinghast, who has written almost exclusively about this topic starting with his second piece.

This week I’d like to continue the discussion with one more alternative to the common metrics that are converted into a saleable inventory structure. I’d like to talk about an idea for a unit called the “session.”


I will start off by saying that this is definitely a more difficult proposition than other kinds of inventory scenarios that have been proposed in the industry but something that the industry should start to think about as an alternative to the way that all media has been bought and sold in the past. As the Internet advertising business has always been interested in being more innovative than its offline-advertising predecessors, this is something about which we need to initiate dialogue.

The first time I ever heard anyone suggest buying and selling media this way was when Brian Monahan, president of Inrhythm Marketing, mentioned it. It must have been more than a year ago when he first floated this idea in conversation at an industry event we were both attending. The concept is this:

    If I can determine how long a unique visitor spends on my site, why not break that time up into units and sell media inventory based on that, rather than the fleeting and quickly forgotten “impression”?

In broadcast, though I am really paying for audience, I am actually buying units of time. For television it is, most commonly, the 15-second or 30-second spot that I’m buying. Sometimes, if I’m a big advertiser (or DR running at 3 a.m.), I’ll buy a longer unit of time.

For radio, the spot lengths are typically 60 seconds for local radio and 30 seconds for network radio.

And, like I mentioned, though I am buying a certain volume of advertising, I’m really paying for a unit of time.

So often, sites love to talk about how long visitors stay. Gaming sites love to brag that the average visitor is there for umpteen minutes per day. Media Metrix and NetRatings both report on average duration of a visit. But what does any of this mean if there is a different ad popping up on the page each time a unique visitor clicks on something?

The length of time a user spends at a site is meaningless to an advertiser buying impressions because what determines the likelihood of a visitor seeing your ad is more a function of the rest of the activity of the site, the number of pages visited, and the placements of your creative. The chance of an advertiser’s ad being seen is only as long as it takes for a unique visitor to click to his or her next location.

So why not sell units of time for online media? Sure, the way it gets priced will have to change, but I don’t think it will change by too much. Truly, an impression is so fleeting in the online space that it’s hardly worth much at all except for a strict direct-response buy. But in the real awareness/branding world, it takes more than a second to make an impression, if you will.

Impressions in broadcast are for as long as the ad unit; if it is a 15-second ad I run, then it is a 15-second impression. Sure, there will be fewer impressions for publishers to sell, but let’s be realistic. They are already selling phantom inventory anyway by counting impressions as pageviews rather than ad views.

By holding publishers to a standard that allows an opportunity for an advertisement to have IMPACT, we might get closer to having real ad models that traditional offline advertisers can accept. And the medium might just start to grow up.

Related reading