It seems everyone’s talking about advertising in RSS (define). It’s no wonder. Consumer adoption is growing. Publishers are increasingly willing to publish RSS feeds. Yet no one’s figured out exactly how to pay for those feeds, much less make money on them. Thankfully, there’s plenty of experimentation afoot.
Feedburner is experimenting widely in RSS advertising. The company, which provides publishers with metrics on subscribers and their behavior, has partnered with almost everyone who’s inserting ads into feeds. Though Feedburner’s coy about some of its relationships, contextual ads from Amazon.com, Yahoo Search Marketing (formerly Overture), Google and Kanoodle have been spotted in feeds they generate.
I sat down recently with Feedburner CEO Dick Costolo to discuss these experiments. In our conversation, he went public for the first time about what the company, its publishers, the ad networks and their advertisers have been learning along the way.
Sometimes Better, Sometimes Worse
The big question on everyone’s mind: how do ads in RSS perform? The first learning Dick shared is ads in RSS perform very differently from ads on the corresponding site — even on the exact same content. Are they better or worse? As in so many cases, the answer is “it depends.”
“Some were much better, some much worse,” Dick told me.
What it comes down to, apparently, is common sense and audience. He shared a couple of examples — both of which describe feeds with over 5,000 subscribers.
In the first case, site “A” displays contextual ads that have an effective CPM of $15. The accompanying feed has CPA ads with an effective CPM of $1.50. Why the disparity?
It turns out the ads on the site are geared toward an audience of people who are discovering the content — product information — via a search engine. Because these folks are at a certain, and very attractive, stage in the buying cycle, advertisers are willing to pay higher prices. Its feed subscribers, on the other hand, may or may not be in the same stage of the buying cycle — they just happen to be interested in that category of product.
“Generally speaking, they aren’t going to be as interested in impulse buying of those products,” Dick said.
In a second situation, site “B” displays contextual ads with an effective CPM of only $1.25, while the RSS feed had ads that resulted in an effective CPM of $6. What happened?
In this case, the difference may be one of targeting. The content in question is tailored toward a design-savvy audience. While the ads on the site were contextually targeted, the ads on the feed were targeted more thematically toward a subscriber base of designers.
Ads Psychologically Connected to Individual Posts
How RSS ads should best be targeted was something else Feedburner considered in its testing. It found readers psychologically associated an ad with the individual post in which it appeared. On a site, the ad is usually considered part of the site as a whole.
On RSS, Dick said, “people are paying much closer attention to the relevance of this particular ad to this specific post.” He added Feedburner reached this conclusion by listening to publisher feedback.
This is a big challenge for many current contextual targeting technologies. Posts are often too short to provide enough information about the proper context. One answer may be to use the overall site content, rather than individual post content, as a basis for targeting.
E-Mail Woes, Multiplied
In its testing, Feedburner got a hint of the creative difficulties ahead for RSS advertising. Dick notes that when the company began doing trials last October, it tracked over 800 different feed aggregators. That number has grown to 1,800, including mobile aggregators, podcast-specific clients and readers used only in certain parts of the world.
In such an environment, how do you move beyond text-only ads and ensure your ad is displayed as you intend? Sure, you can design different creative for different clients, but can you imagine 1,800 different creative executions?
“It’s like the email world with 10 times the complexity,” Dick noted. “For a while, it will be important to provide different form factors based upon the user agent.”
He added that RSS doesn’t even have any format control built in. It was designed more as a data-transfer vehicle rather than as something that would be displayed by itself.
More consolidation and adoption of Web-based readers will rectify this situation somewhat, but I imagine mobile, especially, will continue to confound.
Click-Throughs Going Down
One general theme from Feedburner’s testing will be of special interest to publishers: click-through rates from RSS feeds back to sites are decreasing. The company says this is happening across all the feeds it manages.
“If RSS popularity continues to increase, and it becomes less and less a vehicle for driving site traffic but more and more its own content-viewing medium, that presents an interesting situation to publishers,” Dick said.
He admitted, though, the company didn’t look at full-text versus partial-text feeds. It could be readers are becoming more reluctant to click out of the feed reader, but it could also be that more publishers are providing full-text feeds, giving people fewer reasons to click away.
Whatever the reason, it’s certainly a sign the industry needs to learn lessons like these. If people are staying in RSS readers, advertisers must follow those eyeballs. I’d love to hear from others about what they’ve learned from their own RSS advertising experiences.
Retailers understand the importance and potential of omnichannel marketing, but implementing it is the hard part.
Despite not being one of the juggernauts, Avocados From Mexico made a big impression during the Super Bowl. Meet Ivonne Kinser, who heads the ... read more
While CTRs may have worked in the 1990s, and still do have a place in email marketing, when it comes to banner ads, they’re not your friends when it comes to measuring ad effectiveness. But what other options do we have?
The past month has been filled with big management changes at Twitter, Taco Bell, PayPal, Havas Worldwide, DigitasLBi and Google.