MinnPost, the non-profit news startup in Minneapolis, has rolled out a new form of advertising that looks a little bit like print classifieds, a lot like Twitter, and nothing like traditional marketing on the Internet. They’re calling the service Real-Time Ads, and it’s live in the left column of the front page right now.

The service aggregates tweets, blog posts, and other feeds from local business with timely messages to convey — an ice cream shop announcing the flavor of the day, for instance, or a clothing store offering a one-day coupon. Joel Kramer, the editor and CEO of MinnPost, explains more about about the project in the video above, which I recorded at the Knight Foundation’s conference last week. You can also read the explanation he posted today, which credits local news sites The Deets (in Minneapolis) and The Windy Citizen (in Chicago) with pioneering the idea.

Real-Time Ads could appeal to marketers that already have Twitter accounts and blogs where they’re reaching out to customers, and MinnPost is hoping that the feature will become a destination site the way the classifieds section used to be treasured by newspaper readers. To me, the concept sounds like a viable way for local news sites to finally compete with Craigslist.

In a phone conversation this morning, Kramer described the project as “an effort to move beyond banner ads,” which have proven lucrative for the site, but also limited. MinnPost charges, on average, $15 per thousand impressions for its display advertising, but because the ads are for local businesses, they’re only served to readers in Minnesota. (That’s one reason the CPM is so good.) As a result, the site had several weeks this spring when marketers’ demand for advertising space exceeded the site’s supply of local readership.

That’s an envious dilemma, but it also means that MinnPost hasn’t maximized its advertising as it seeks to wean off foundation support. Kramer is hoping that Real-Time Ads will provide a new source of revenue “that isn’t as tied to traffic as banner ads are.” MinnPost is offering the space for free at the start but plans to charge “under $100 a week,” a price point that’s close to print classifieds ads in their heyday.

MinnPost, like most publishers, has struggled to provide an option for local advertisers with small budgets that can’t accommodate banner ads. They’ve been using Flyerboard, an ad platform by New Haven startup PaperG that’s popular among local news sites, but Kramer told me: “It has not produced a lot of business for us. We do believe Real-Time Ads will prove more valuable for advertising at a lower entry point.”

There’s more detail in the video above, including this bit of philosophy from Kramer:

It’s an experiment. We’ve tried many experiments. Some of them work, some of them don’t.  We don’t get too worried about the ones that don’t. And I’m a big believer in that you have to try a lot of things and get a few victories.

And here’s a full transcript of the video:

Joel Kramer: We’re launching a beta test of a kind of advertising that’s different from the banner ads that have been our main source of revenue in advertising. We have banner ads, and we have sponsorships that are a kind of a banner component and other kinds of advertising to them. But we decided it was important to try some other things that maybe took better advantage of what the Internet makes possible, and one thing we got very intrigued by recently was watching marketers — both for-profits and non-profit organizations and governments, people who want to communicate — watching them use Twitter and their own RSS feeds to communicate to their audiences. It’s a growing phenomenon that people are doing this, and one thing it has is an element of immediacy to it. In other words, the marketer can communicate several times even the same day, whereas a banner ad is likely to be changed out every couple of weeks.

So we’re creating a beta test, called MinnPost Real-Time Ads, in which we will take marketers, ranging from non-profits to government agencies to companies selling things, who have RSS feeds, maybe they’ve been marketing on Twitter or through other RSS feeds, and give them a space on our site, in which they can update their own messages frequently and create a place on the site where the reader can know that they can go to get ads with a great deal of immediacy. That they’ll get the latest messages from a lot of different marketers. And we’re going to try it by giving it away free, for a while, to marketers in the community that we can see are already doing it because we’re reading their feeds. […]

It’s all local. You know, we appeal to a local audience and we are looking for local advertisers. The national advertising business pays so little, on a CPM basis now, that its not really very promising. So this is all local. And the test will be done on a small scale, you know maybe somewhere between 8 and 16 or so advertisers. But this could be done on a very big scale. You could think it could become kind of like a new version of classified ads in which many, many players at a low price point can communicate regularly to our audience. […]

We believe this kind of advertising is form of content, and, for example, several of the ads might on a rotating basis might appear on the regular pages of the site and you can click through to see more of them. But, for example, if the Department of Transportation is putting out an RSS feed about construction tie-ups around the city, and if they would pay a small amount of money to put that on our site, that would be interesting information to our readers. Or whether there are parking spaces available at the airport or what a restaurant has on its menu as a special tonight for dinner. These are the kinds of things that could change everyday and actually would be, I think, of interest to readers. […]

One limiting factor is that not that many yet have good RSS feeds, and we have to train some people to do it. But at the beginning we are going to try to work with the leaders who are already out there, and they’ll show other people how to do it. […]

The advantage is, you know, every marketer has its own list, so they could say, well, we don’t need you, we’re already doing this on Twitter, or we have our own RSS feed. But the advantage is for them to reach new people because we would be creating a kind of marketplace of immediacy. We would be creating a place where readers who want to see a lot of different kinds of recent messages can go. And, if we had enough marketers we could start to classify the ads by categories. For example, if we had a lot of restaurants, well, you could look  at only the restaurants and see what the restaurants have for dinner today.  And then if we had even more, you could break it down by restaurants in different parts of the community.

So we’re building the software that will enable us to do this. What we add to the marketers’ a new audience. They already have the audience they have, but without creating any new content, we can give them a chance to send their message to more people. […]

It’ll be on an out-of-pocket basis, clearly a lot cheaper than banner advertising, and that opens up the opportunity for us to reach people who don’t want to pay as much as banner ads cost. But it won’t be sold on a CPM basis because much of the use of it will be on advertising-only pages. So, in that respect, it’s more like classified in the old days when, you know, readers would go to a section and actually look at the ads because they were looking for it as content. Well, here, the amount of traffic we have on the site will not be the defining factor. The issue will be how many people want to go look for these ads. […]

It’s an experiment. We’ve tried many experiments. Some of them work, some of them don’t.  We don’t get too worried about the ones that don’t. And I’m a big believer in that you have to try a lot of things and get a few victories. […]

Our technology director, Karl Pearson-Cater, worked with me on the basic idea, and he has used two outside people to help build it. And, there is some work that goes into that, and you know if it works well we might be able to get some benefit by sharing that with other people who want to use it. Yeah. Either, one way or another, either selling software or rent, or providing a service to people. But that’s down the road, we have to first see if it works.