There’s an interesting conversation on new ad models going on in the Poynter online news e-mail list that started yesterday with Steve Yelvington, who forwarded a post from his blog for discussion:

The argument goes like this: We have an audience problem. We can fix our sales incentives, train our people, tune our pricing and our packaging, and replace leadership as necessary. But at the end of the day we’re going to hit a very hard wall. That wall is /available advertising inventory/ that meets the advertisers’ needs. That inventory comes from audience, from reach (unique users) multiplied by frequency (pageviews per user). And while the reach numbers may look good, the frequency numbers suck.

Greg Harmon from Belden Associates chimed in to break down reader frequency into three groups and their mean frequency of visits:

• Fly-by: Driven by episodic events picked up by Drudge, Digg, Yahoo, Google, whatever. Highly variable and can range from 15-20% to over 60% of monthly site traffic. Mean frequency: One visit per month.
• Loyalist Incidentals: Familiar with the site but infrequent and come for a specific purpose a few days per month for a story, weather, etc. Inconclusive data makes it hard to separate this group from fly-bys. May be 40% of traffic.Mean frequency: Two times per month.
• Core Loyalists: Most important; the real audience for advertisers. Visit an average four days per week, two to three times on weekdays. About 40% of traffic. Mean frequency: 40 times per month.

Harmon goes on to say that cookie-based reporting tools are part of our problem because people reset them, etc., making numbers are suspect.
Vin Crosby rode in with numbers from the NAA and Nielsen//Netratings data from March through August 2007:

She (The site’s media kit says that 57.1 percent of users are female) visited only 4.05 times per month, saw only 27 pages per month, and spent a total of 20 minutes and 20 seconds on the site during all those visits. * This means the AVERAGE USER of the premier DAILY newspaper in the U.S. visits LESS THAN ONCE PER WEEK.
* It means that she probably reads LESS THAN 27 stories there all month (because NYTimes.com, the maximize its number of banner ad exposures, tends to stretch stories over more than one Web page each).
* It means that she probably spends LESS THAN FIVE MINUTES on the site in each of those infrequent visit.
* And it also means that the NYTimes.com’s average user probably reads less of that newspaper’s content in a month than the paper’s average print edition readers reads in a single day.

His conclusion:

You can claim that daily newspapers (whether in print or online) nowadays reach an ‘exclusive audience,’ but the plain fact is that newspapers are general-interest publications. You can claim that frequency doesn’t matter, but I think that advertisers (and the newspaper sites’ ad sales people’s pitches) will say different. The bottom line is that 12 years after American daily newspapers first began publishing on the Web, their sites are read FAR LESS OFTEN AND FAR LESS THOROUGHLY than their companies’ dying newsprint editions and earn only a SMALL FRACTION of those dying editions’ revenues.

Good numbers will certainly help flesh out the picture, but it only serves to drive home the point in my mind that our focus in the newsroom must shift from the print publication to what we’re able to do online.

I’m still struggling to decipher Omniture data on my reporting blog, but I seem to have the same core readers. My paper is doing a self-promo featuring reporters out working with the tagline “More reporters, more local news.” You get the picture. Problem is, it’s made up of ads in the paper–preaching to the converted.

First thing is to get reporters aggressively (and effectively) blogging to create groupies and enhance the kind of content that only lives online. Then get the ads promoting the newsroom in other places: movie theaters, TV, billboards.

As a reporter, I’m a bit uncomfortable being the center of attention, but these are desperate times.

Thoughts on this? New ad models?